Offset correction of angular position sensor in an electric machine

Abstract

The invention relates to a machine (12) and to a method for operating an electric machine (12). The electric machine (12) comprises a rotor (18). An angular position sensor (24) and an evaluation unit (30) are associated with the electric machine (12). The rotor (18) is arranged according to a rotor angle and rotates according to a rotational speed. The angular position sensor (24) is designed to detect a first and a second output signal relating to the rotor angle and to transmit same to the evaluation unit (30). The evaluation unit (30) is designed to determine offsets between a first and a second angular component of the rotor angle and a reference angle on the basis of at least one low-pass filter (34). The evaluation unit (30) is also designed to determine amplitudes of the first and a second angular component of the rotor angle on the basis of a Fourier transform for a single support frequency. The support frequency corresponds to the rotational speed of the rotor (18).

Description

[0001] Electrical machine and method for operating an electrical machine

[0002] An electric machine and a method for operating an electric machine are provided.

[0003] For the operation of electrical machines, particularly to ensure they deliver the desired torque, it is necessary to precisely determine the rotor angle in order to accurately regulate the current. This is typically achieved using angle position sensors that provide a first and a second measurement signal. These signals should be orthogonal to each other according to a Cartesian coordinate system and together represent an amplitude that moves around a unit circle at the coordinate center. From these two signals, the angle at which the amplitude is oriented at each control point can be determined. This angle corresponds to the rotor angle at that control point. The rotor speed can then be calculated from the sequence of rotor angles over several control periods.

[0004] In general, the measurement signals are subject to interference. Therefore, the measurement signals exhibit errors corresponding to a limited measurement accuracy. For example, manufacturing tolerances or material defects in the electric motor and / or the angle position sensor can cause interference, which is why the amplitude of the measurement signals can generally vary. Furthermore, interference can also lead to the individual measurement signals not being orthogonally oriented to each other, but rather exhibiting offsets from a reference angle. Therefore, the individual measurement signals are normalized to a predefined value. Finally, the rotor angle for the respective control period can then be determined, which can be used for further control.

[0005] A common approach starts with a unit circle, around whose circumference the total amplitude based on both measurement signals rotates around the coordinate center. Mathematically, the measurement signals can be described using sin. 2+ cos 2 The sine and cosine signals must be related to each other to normalize the partial signals. The sine and cosine signals reflect the individual measurement signals of the angle position sensor. This means that the individual measurement signals must be normalized against the amplitude of the unit circle (corresponding to the radius r = 1), which, however, is generally variable due to interference.

[0006] In order to ensure the functionality of the control system of the electric machine, for example for a desired ASIL (ASIL: “Automotive safety integrity level”), the measurement signals are therefore usually checked for plausibility.

[0007] Conventional plausibility checks are based on a unit circle with a predefined, undisturbed (perfect) circular shape and a predefined amplitude. The measurement signals are then simulated using modified normalization factors until they match the unit circle according to the relationship mentioned above. The signal amplitude for each control period can be determined using these normalization factors, allowing the individual measurement signals to be validated against each other and / or against the overall amplitude.

[0008] The simulation used for plausibility checks causes a high level of control effort, which is computationally intensive and means that the corresponding computing devices are occupied for a significant portion of the respective control period, and are therefore not available for other control aspects.

[0009] Furthermore, this approach leads to the fact that, in principle, the same method is used to determine the rotor angle for verifying the plausibility of the individual measurement signals. This is generally undesirable for plausibility checks, in order to rule out the possibility that systematic errors in both calculation methods have the same effect and therefore potentially remain undetected.

[0010] Therefore, one task is to provide an electrical machine for which the measurement signals can be validated in a less computationally intensive way and based on a different approach.

[0011] The problem is solved according to the invention by an electric machine with a rotor to which an angular position sensor and an evaluation unit are assigned. The rotor is arranged according to a rotor angle and rotates at a speed corresponding to a rotational speed. The angular position sensor is configured to acquire a first and a second output signal with respect to the rotor angle and transmit them to the evaluation unit. The evaluation unit is configured to determine offsets between a first and a second angular component of the rotor angle and a reference angle based on at least one low-pass filter. The evaluation unit is further configured to determine the amplitudes of the first and the second angular component of the rotor angle by means of a Fourier transform for a single support frequency. The support frequency corresponds to the rotational speed of the rotor.

[0012] The invention is based on the understanding that the rotational speed of the rotor changes only slowly compared to its angular position. Therefore, the output signals of the angular position sensor can be evaluated for a Fourier transform based on a single reference frequency to determine the amplitudes for that specific frequency. All other frequency-dependent amplitudes can be neglected. Simultaneously, the respective offsets between the angular components and the reference angle can be determined using a computationally efficient low-pass filter. This allows the rotor angle to be determined precisely and with minimal effort.

[0013] This approach is completely independent of the previous approach using the unit circle. Therefore, the values ​​determined using the new approach can be used to perform a computationally efficient, efficient, and reliable plausibility check on previously determined corresponding values ​​obtained using a different approach (with the unit circle). This prevents systematic errors from going undetected during the plausibility check, as two independent procedures can be used.

[0014] Optionally, the evaluation unit includes a data memory in which it stores determined offsets and / or amplitudes. This allows the evaluation unit to take into account the determined offsets and / or amplitudes from previous control periods. Since the rotational speed, for example, can also be determined from the offsets and amplitudes, the rotational speed of a previous control period can be used for a subsequent control period, preferably immediately, to define the support frequency for which the Fourier transform is performed.

[0015] Preferably, the angular position sensor has at least two non-parallel oriented measuring coils. Each output signal is based on one measuring coil. Thus, the output signals are acquired and provided by independent measuring elements. The measuring coils are oriented relative to each other according to a predefined geometric angle, for example, orthogonally to each other. This allows the measuring coils to represent different angular components of the rotor angle, which can be mapped onto the measuring coils, for example, by a rotor coil of the angular position sensor.

[0016] In one aspect, the evaluation unit can at least include a lookup table containing different time constants of the low-pass filter depending on the rotor speed. This allows several rotor speed ranges to be differentiated with regard to the low-pass filter's behavior. The time constant of a low-pass filter is inversely proportional to its cutoff frequency. The larger the time constant of the low-pass filter, the smaller the amplitude of its output signal for a given, constant input frequency.

[0017] The output signals of the angle position sensor are frequency-dependent and can, for example, exhibit a sinusoidal profile. If an unsuitable time constant is selected for the low-pass filter for a specific speed range, the time required to determine the offsets between the angular components of the rotor angle and the reference angle becomes undesirably long. This, in turn, would block the processing units used in the evaluation process. However, the lookup table allows for the selection of a suitable time constant for the low-pass filter for the respective rotor speed range, thus optimizing the time required to determine the offsets.

[0018] Optionally, the low-pass filter is multifunctional and can be operated according to different time constants. The low-pass filter can be controlled by the evaluation unit or a connected device, such as a control unit, depending on the respective time constant. For example, control signals can be used to close and / or open switching devices, which allows additional components of the low-pass filter to be added, such as inductors, resistors, and / or capacitors.

[0019] Alternatively, different low-pass filters can be used to determine the offsets for different time constants.

[0020] Preferably, the angular position sensor is configured as a transmitter, resolver, or coordinate converter. This allows the angular position of the electric machine's rotor to be represented as one or more electrical signals. The output signals of the angular position sensor depend directly on the rotor's angle of rotation.

[0021] Generally, the angular position sensor can be separate from the electric machine. Alternatively, the angular position sensor can also be considered part of the electric machine and form a stationary unit with it.

[0022] The electric machine can be used in a vehicle, for example as a drive motor.

[0023] According to a further aspect, the problem is also solved according to the invention by a method for operating an electric machine with a rotor that is arranged according to a rotor angle and rotates according to a speed. The electric machine is assigned at least one angular position sensor and an evaluation unit. The method comprises at least the following steps:

[0024] The evaluation unit receives a first and a second output signal from the angular position sensor, which is assigned to the rotor.

[0025] Offsets between a first and a second angular component of the rotor angle and a reference angle are determined by the evaluation unit based on at least one low-pass filter.

[0026] The evaluation unit determines the amplitudes of the first and a second angular component of the rotor angle using a Fourier transform for a single reference frequency. The reference frequency corresponds to the rotor speed. This method allows the amplitudes of the angular components and their offsets with respect to a reference angle to be determined independently of previous state-of-the-art methods based on the unit circle. Since the rotor's speed and angle can be precisely characterized by the determined variables, previously determined corresponding values ​​can be validated. This can, for example, ensure a higher level of functional reliability, such as ASIL-C. It should be noted that the same level of precision is not required as is advantageous in current control for reducing power losses.However, it is important that values ​​determined according to conventional methods, which characterize the rotor condition, can be assessed qualitatively and quantitatively using a different method in order to exclude systematic errors.

[0027] Preferably, the evaluation unit takes into account the rotor speed from a previous control period in a subsequent control period, preferably immediately. For this purpose, a data storage device can be used, for example, in which the speed values ​​are stored. Compared to a change in the orientation of the rotor's magnetic field, the rotor speed changes only slowly. Therefore, this approach reduces the complexity.

[0028] Optionally, the first and second output signals of the angle position sensor are differential signals with respect to a previous control period. This means that the evaluation unit only considers the changes in the output signals for successive control periods. As a result, any offsets between the angular components of the rotor angle and the reference angle are small in magnitude and exhibit only minimal variation due to the differential measurement method. In other words, the offsets remain approximately constant for successive control periods because they change only slowly. This can also be used for plausibility checks, since the range of values ​​expected for the offsets in the subsequent control period is known based on the previous control period.According to one aspect, the evaluation unit can determine the rotor speed based on its clock rate and the first and / or second output signal of the angle position sensor. The rotational speed corresponds to the angle swept per unit of time. The unit of time is defined by the clock rate of the evaluation unit. The swept angle corresponds to the change in the rotor angle, which is determined for a predetermined number of control periods. This allows for efficient determination of the rotational speed due to the differential acquisition of the output signals.

[0029] In some embodiments, the evaluation unit can take into account different time constants of the low-pass filter depending on the rotor speed. For this purpose, a lookup table can be used, in which speed ranges are distinguished with respect to the rotor speed by corresponding limit values. In this way, the time constant can be selected as a function of the rotor speed in such a way that the time required to determine the offsets between the angular components of the rotor angle and the reference angle can be optimized.

[0030] Preferably, the determined offsets and / or amplitudes are stored by the evaluation unit in a data memory. This allows the corresponding values ​​to be used for subsequent control periods. For example, this allows successive values ​​of the determined offsets and / or amplitudes of the angular components of the rotor angle to be validated based on the preceding values.

[0031] Optionally, the evaluation unit determines offset differences between the calculated offsets and / or amplitude differences between the calculated amplitudes of successive control periods. The evaluation unit outputs the calculated offsets and / or calculated amplitudes for an external component only if the offset differences are less than a first differential threshold and the amplitude differences are less than a second differential threshold. The differential thresholds are thus used as a plausibility criterion to prevent highly fluctuating values ​​from other components from being used to control the electric machine, for example, in current control. Typically, the values ​​of the calculated offsets and / or amplitudes of the angular components of the rotor angle change only slowly for successive control periods.This is because the control frequency is significantly higher than the frequency corresponding to the rotor speed. In other words, the rotor speed can be assumed to be constant for several control periods. Therefore, differential thresholds can be defined, which are maintained by the corresponding differences during fault-free operation of the electrical machine. If the differential thresholds are exceeded, this indicates a fault. To prevent the fault from propagating to the control system implemented by other components, the output of the determined offsets and / or amplitudes can be suppressed in this case. Simultaneously, a fault notification can be sent to a user interface and / or an external component.

[0032] According to one aspect, the method can be based on certain assumptions. For example, it can be assumed to a good approximation that the output signals of the angle position sensor have a sinusoidal waveform. Furthermore, the range of values ​​within which the amplitudes of the angular components of the rotor angle are expected in a specific control period is known, at least to a limited extent. For instance, the values ​​from the previous control period can be taken into account.

[0033] The amplitudes of the angular components of the rotor angle Amplitude (SIN} and Amplitude (C OS) can be determined in particular based on a Fourier transform for a single support frequency 9:

[0034] Amplitude (SIN} = yjLPF(sm(9} ■ sin raw} 2 + LPF(cos (9} ■ sin raw} 2

[0035] A

[0036]

[0037] amplitude (C OS} = ^LPF(sm(9} ■ cos raw} 2+ LPF(cos (9} ■ cos raw} 2 In this context, sin refers to raw and cos raw The unchanged output signals of the angle position sensor. LPF denotes the low-pass filter, which has a speed-dependent time constant. This allows the amplitudes of the angular components of the rotor angle to be determined with particularly high computational efficiency. The calculation and evaluation steps performed by the evaluation unit can also be based on a computer program that contains instructions which, when executed by a processor, cause the processor to perform the corresponding calculation and evaluation steps.

[0038] According to a further aspect, the problem is also solved by a drive unit comprising an electric machine, an inverter, a control device, and an angular position sensor. The electric machine is designed as described herein. The inverter is associated with the electric machine. The control device comprises the evaluation unit.

[0039] The vehicle can be a land vehicle, rail vehicle, aircraft, or watercraft. It can be a hybrid vehicle, which has both an internal combustion engine and an electric motor, or it can be an electric vehicle.

[0040] The invention, as well as further advantageous embodiments and developments thereof, are described and explained in more detail below with reference to the examples shown in the drawings. The drawings show:

[0041] Fig. 1 shows a schematic representation of a drive unit with an electric machine according to the invention.

[0042] Fig. 2 shows a schematic representation of a method according to the invention for operating an electric machine, and

[0043] Fig. 3 shows a schematic representation of a method for normalizing the output signals of the angular position sensor.

[0044] All features mentioned below with reference to the exemplary embodiments and / or the accompanying figures can be combined alone or in any subcombination with features of the invention, including features of preferred embodiments.

[0045] Fig. 1 shows a schematic representation of a drive unit 10 with an electric machine 12 according to the invention. The drive unit 10 comprises a converter 14, which is associated with the electric machine 12 and provides phase currents for the windings L1, L2, L3 of the stator 16 so that the rotor 18 can be driven. In the normal operating state of the electric machine 12, the phase currents are regulated such that a desired torque is output via the rotor 18 to an external component.

[0046] According to this embodiment, the electric machine 12 is a separately excited electric machine 12, such that the rotor 18 has a rotor winding 20 which is supplied with a rotor current by means of a rotor circuit 22 in order to generate a magnetic field.

[0047] The drive unit 10 also includes an angular position sensor 24, which is associated with the electric machine 12. According to this embodiment, the angular position sensor 24 is configured as a resolver. The angular position sensor 24 itself has a rotor coil 26, which is coupled to the rotor 18 of the electric machine 12. Furthermore, the angular position sensor 24 has two measuring coils 28, which are not oriented parallel to each other. Since the rotor coil 26 of the angular position sensor 24 rotates in accordance with the rotation of the rotor 18 itself, corresponding to the rotor angle 9, varying voltages are induced in the measuring coils 28 of the angular position sensor 24. This provides a time-varying output signal for an evaluation unit 30 via each measuring coil 28.

[0048] The evaluation unit 30 comprises at least a data processing device 32, a low-pass filter 34, and a data storage device 36. The data storage device 36 contains at least lookup tables 38 for the low-pass filter 34. The lookup tables 38 specify different time constants of the low-pass filter 34 as a function of the rotational speed of the rotor 18 of the electric machine 12.

[0049] The evaluation unit 30 is also coupled with a user interface 40, via which the evaluation unit 30 can output notifications, for example error notifications.

[0050] Furthermore, the evaluation unit 30 is coupled to a control device 42, which is assigned to the inverter 14 of the drive unit 10. This allows parameters determined by the evaluation unit 30 to be transmitted to the control device 42 and taken into account by the latter when controlling the inverter 14.

[0051] Alternatively, the evaluation unit 30 and the control device 42 can also be integrally designed together as a single component.

[0052] Figure 2 schematically illustrates a method according to the invention for operating an electric machine 12. Optional steps are shown with dashed lines.

[0053] According to optional step S2, the output signals of the different measuring coils 28 of the angular position sensor 24 are acquired differentially. This means that only the change in the output signal compared to the previous control period of the evaluation unit 30 is acquired. As a result, disturbances that change only slowly relative to the frequency of the control period can be disregarded, since they have only a negligible influence due to the differential acquisition method.

[0054] In step S4, a first and a second output signal from the angular position sensor 24, which is assigned to the rotor 18, are received by the evaluation unit 30. For this purpose, the evaluation unit 30 can, for example, have dedicated signal inputs.

[0055] In the subsequent step S6, offsets between a first and a second angular component of the rotor angle and a reference angle are determined by the evaluation unit 30 based on the low-pass filter 34. This means that the system evaluates the extent to which the angular components are non-orthogonal to each other and how much the offsets have changed compared to the previous control period. Under normal operating conditions of the electric machine 12, the offsets are negligibly small due to the differential measurement method.

[0056] Step S6 can be further developed by the optional steps S8 to S12. In optional step S8, the evaluation unit 30 takes into account the rotational speed of the rotor 18 from a previous control period. This utilizes the fact that the rotational speed of the rotor 18 changes only slowly compared to the clock frequency of the evaluation unit 30. The rotational speed of the rotor 18 can therefore be considered constant over several control periods. Knowing the rotational speed of the rotor 18, the evaluation unit 30 selects a time constant for the low-pass filter 34 from the lookup table 38 of the data memory 36, corresponding to optional step S10. This time constant must be taken into account when determining the offsets. This allows the time required to determine the offsets to be optimized.

[0057] According to optional step S12, the low-pass filter 34 is set by the evaluation unit 30 according to the read-out time constant. The offsets are then determined according to step S6.

[0058] Following step S6, the evaluation unit 30 determines the amplitudes of the first and a second angular component of the rotor angle in the subsequent step S14 using a Fourier transform for a single support frequency. The support frequency corresponds to the rotational speed of the rotor 18.

[0059] In general, the output signals of the measuring coils 28 can be transformed into the frequency domain using a Fourier transform, allowing the amplitudes for different frequencies to be determined. However, since the rotational speed of the rotor 18 can be assumed to be constant for several control periods, the amplitudes for frequencies that do not correspond to the rotational speed of the rotor 18 can be neglected. It is assumed that the output signals of the measuring coils 28 of the angular position sensor 24 have a sinusoidal waveform and that the range of values ​​in which the amplitudes of the angular components of the rotor angle of the rotor 18 are expected to occur in a control period is known to a limited extent. In other words, based on previous control periods and knowledge of the rotational speed of the rotor 18, the approximate values ​​that the amplitudes of the angular components of the rotor angle of the rotor 18 should have are generally known.Therefore, the rotational speed of rotor 18 from the previous control period can be used as the support frequency.

[0060] The procedure can then be further developed by the optional step S16, in which the evaluation unit 30 takes its evaluation clock frequency into account to determine the rotational speed of the rotor 18. The rotational speed of the rotor 18 can be determined based on the rotor angle swept out per unit of time. The time unit is defined by the evaluation clock frequency. Due to the differential acquisition method of the output signals of the measuring coils 28, only the amplitude change that corresponds to the angular change of the rotor angle compared to the previous control period is recorded. In this way, the rotational speed of the rotor 18 can be determined efficiently.

[0061] Furthermore, the procedure can be extended by the optional step S18, in which the determined offsets and / or amplitudes of the angular components of the rotor angle of rotor 18 are stored in the data memory 36 of the evaluation unit 30. This allows them to be used for subsequent control periods.

[0062] Optionally, the evaluation unit 30 can, according to step S20, consider several determined offsets and / or several determined amplitudes of the angular components from different control periods and determine offset differences and / or amplitude differences. These offset differences and / or amplitude differences can then be compared with a respective difference threshold value to verify whether the determined offsets and / or amplitudes correspond to the expected operating behavior of the electric machine 12 or whether they are faulty. In particular, this allows determination of whether the offsets and / or amplitudes exhibit strong fluctuations or not. This is equivalent to stable or unstable operation of the electric machine 12.

[0063] Based on the comparison from optional step S20, the determined offsets and / or amplitudes can be provided by the evaluation unit 30 for external components, for example for the control device 42, according to optional step S22. In general, the determined offsets and / or amplitudes can then be used for normalization, for example in subsequent control periods.

[0064] If the differential thresholds are exceeded in optional step S20, the evaluation unit 30 can issue an error message according to optional step S24, for example via the user interface 40. In this case, the output of the determined offsets and / or amplitudes for external components for the purpose of controlling the electric machines 12 is prevented. The method described here serves to validate the offsets and / or amplitudes of the angular components of the rotor angle of the rotor 18 and / or the rotational speed of the rotor 18. Generally, these parameters are also determined within the framework of controlling the electric machine 12 by other devices, for example the control device 42, based on a different approach, for example using the unit circle.In contrast, the method presented here uses a Fourier transform for a single support frequency, allowing the determined values ​​of the variables to be compared with one another to enable a robust plausibility check. This increases the functional reliability of the control of the electric machine 12. The use of the Fourier transform has the advantage over previous methods that interference effects caused by harmonics are essentially suppressed. Even if the plausibility check does not require the same level of precision that is desirable in the context of current control for the electric machine 12, the precision can still be improved compared to previous approaches due to the suppression of interference effects from harmonics.

[0065] Fig. 3 shows a schematic representation of an exemplary method for normalizing the output signals of the angular position sensor 24.

[0066] First, in step S30, it is determined whether the rotational speed of rotor 18 is less than a predefined speed limit. The rotational speed from a previous control period can be taken into account for this purpose.

[0067] If this is the case, the following step S32 determines whether "stable" values ​​for the offsets and / or amplitudes have been obtained, i.e., values ​​that exhibit only minor changes over time. In general, the offsets and / or amplitudes have not been determined using the method shown in Fig. 2. For example, they may be values ​​determined using the unit circle according to previous methods.

[0068] However, corresponding to the optional step S20 from Fig. 2, offset differences and / or amplitude differences can be determined and compared with corresponding difference thresholds. If the difference thresholds are not exceeded, it can be assumed that the determined offsets and / or amplitudes of the angular components of the rotor angle of rotor 18 are stable.

[0069] If, in step S32, it is determined that the calculated offsets and / or amplitudes are unstable, for example, because the offset changes and / or amplitude changes exceed their respective limit values, the output signals of the angular position sensor 24 are normalized in the subsequent step S34 based on the expected values ​​of the amplitudes of the angular components of the rotor angle of the rotor 18. This means that the calculated values ​​of the offsets and / or amplitudes from the control period are disregarded for normalization at this point. Instead, the system takes advantage of the fact that the parameters change only slowly over time, which is why expected parameter ranges for the offsets and / or amplitudes of the angular components are known from previous control periods. These values, determined in previous control periods, are then used for nomination in the subsequent control period.

[0070] If, in step S32, it is determined that the calculated offsets and / or amplitudes are stable, for example because the offset changes and / or amplitude changes are smaller than a respective limit value, the values ​​of the offsets and / or amplitudes determined in this control period are used in accordance with the following step S36 to normalize the output signals of the angular position sensor 24.

[0071] If, alternatively, it is determined in step S30 that the rotational speed of the rotor 18 of the electric machine 12 exceeds the specified rotational speed limit, the offsets and / or amplitudes of the angular components of the rotor angle of the rotor 18 are determined again using the method from Fig. 2 in accordance with step S38 in order to enable a plausibility check.

[0072] In the subsequent step S40, it is determined whether stable values ​​for the offsets and / or amplitudes have been obtained using the method described with reference to Fig. 2. In this step, the optional step S20 of the method can be taken into account. If it is determined in step S40 that no stable values ​​for the offsets and / or amplitudes can be obtained using the method described with reference to Fig. 2, then, according to step S42, the expected values ​​of the amplitudes are used to normalize the output signals of the angular position sensor 24, as described in step S34.

[0073] If, according to the optional step S20 of the procedure from Fig. 2, it is determined that the determined offset changes and / or amplitude changes do not exceed corresponding difference threshold values ​​and are therefore stable, the values ​​of the offsets and / or amplitudes determined on the basis of the procedure described with reference to Fig. 2 are used according to step S44 for normalizing the output signals of the angular position sensor 24.

[0074] This allows the method described with reference to Fig. 2 to be used directly in the control of the electric machine 12. Furthermore, values ​​of offsets and / or amplitudes determined using other methods can be validated.

Claims

Patent claims 1. Electric machine (12) with a rotor (18) to which an angular position sensor (24) and an evaluation unit (30) are assigned, wherein the rotor (18) is arranged according to a rotor angle and rotates according to a rotational speed, wherein the angular position sensor (24) is configured to detect a first and a second output signal with respect to the rotor angle and transmit them to the evaluation unit (30), wherein the evaluation unit (30) is configured to determine offsets between a first and a second angular component of the rotor angle and a reference angle based on at least one low-pass filter (34), and wherein the evaluation unit (30) is further configured to determine amplitudes of the first and the second angular component of the rotor angle by means of a Fourier transform for a single support frequency, and wherein the support frequency corresponds to the rotational speed of the rotor (18).

2. Electric machine (12) according to claim 1, characterized in that the evaluation unit (30) has a data storage device (36) in which the evaluation unit (30) stores determined offsets and / or determined amplitudes.

3. Electric machine (12) according to claim 1 or 2, characterized in that the angular position sensor (24) has at least two non-parallel oriented measuring coils (28), and wherein each output signal is based on a measuring coil (28).

4. Electric machine (12) according to one of the preceding claims, characterized in that the evaluation unit (30) has at least one lookup table which includes different time constants of the low-pass filter (34) depending on the rotational speed of the rotor (18).

5. Method for operating an electric machine (12) with a rotor (18) arranged according to a rotor angle and rotating according to a speed, wherein the electric machine (12) is assigned at least one angular position sensor (24) and an evaluation unit (30), the method comprising the following steps: A first and a second output signal of the angular position sensor (24) assigned to the rotor (18) are received by the evaluation unit (30); Offsets between a first and a second angular component of the rotor angle and a reference angle are determined by the evaluation unit (30) based on at least one low-pass filter (34); The amplitudes of the first and a second angular component of the rotor angle are determined by the evaluation unit (30) using a Fourier transform for a single support frequency, where the support frequency corresponds to the rotational speed of the rotor (18).

6. Method according to claim 5, characterized in that the evaluation unit (30) takes into account the rotational speed of the rotor (18) from a previous control period.

7. Method according to claim 5 or 6, characterized in that the first and the second output signal of the angular position sensor (24) are differential signals with respect to a previous control period.

8. Method according to one of claims 5 to 7, characterized in that the evaluation unit (30) determines the rotational speed of the rotor (18) based on a clock rate of the evaluation unit (30) and the first and / or the second output signal of the angular position sensor (24).

9. Method according to one of claims 5 to 8, characterized in that the evaluation unit (30) takes into account different time constants of the low-pass filter (34) depending on the rotational speed of the rotor (18).

10. Method according to one of claims 5 to 9, characterized in that the determined offsets and / or determined amplitudes are stored by the evaluation unit (30) in a data storage device (36).

11. Method according to one of claims 5 to 10, characterized in that the evaluation unit (30) determines offset differences between the determined offsets and / or amplitude differences between the determined amplitudes of successive control periods, and wherein the evaluation unit (30) outputs the determined offsets and / or determined amplitudes for an external component only if the offset differences are smaller than a first difference threshold and the amplitude differences are smaller than a second difference threshold.

Images