DE102014206048A1 - Method for operating an electric motor - Google Patents

Abstract

In the method according to the invention for operating an electric motor in a system, for example an automobile, a correction signal for changing the operating noise of the system at at least one specifiable frequency is determined from a frequency-dependent response function of the system and is used to control the electric motor. The method allows a control of an electric motor in which a predeterminable setpoint characteristic of the system can be achieved, taking into account the effective paths in the system during sound transmission, for example via structure-borne noise through the system and / or via airborne sound. The response function is in particular the frequency response function, which represents the transmission behavior of the system. In particular, this response function is determined as a function of the operating point. This procedure enables interference noises to be reduced and / or certain target characteristics of the system to be impressed. The correction signal can be determined in advance once per operating point (passive method) and / or be determined continuously or temporarily during operation (active method).

Description

The present invention relates to a method for operating an electric motor that serves the acoustic optimization of electric motors, in particular for motor vehicles.

Although electric motors, which are used as traction machines, basically have a lower noise emission than comparable performance internal combustion engines, the specific noise behavior of electric motors in many applications - especially in the field of electromobility - to comfort impairments to acceptance problems. The reason for this lies mainly in the completely different from the internal combustion engines noise characteristics of electric motors. By way of example, reference is made here to the unsteady level profile and the very tonal, order-specific howling of the electric motors, which is further intensified by the considerably larger speed range of the electric motors (often greater than 10,000 rpm) combined with the high atomic numbers. The noise characteristic of known electric motors is very unusual for car drivers and represents a major hurdle with respect to the acceptance of electric vehicles. The optimization of the noise behavior of electric vehicles is therefore an important task of vehicle manufacturers or suppliers and is considered a major challenge.

The cause of the relatively high-frequency and tonal noise characteristics of electric motors is primarily related to the conceptual change in the magnetic flux, resulting in permanent magnet motors depending on the number of poles and windings to prominent orders. These changes in the magnetic flux generate predominantly radial, but also axial and tangential alternating forces both on the engine components (rotor, stator) and on the motor housing or the coupled structures. The chain of action starting from the magnetic flux change, via the alternating force excitation of the engine components and their structural dynamic vibration behavior (for example, rotor dynamics) leads to vibration excitation of the motor housing. Very important here is the coupling of the electromagnetic excitation with the structural dynamics. This means that an electromagnetic excitation causes a movement of the rotor, for example, but this rotor movement in turn reacts back to the electromagnetic excitation. This circumstance makes the consideration of such an overall system extremely complex with the means of simulation and has hitherto not been considered by any of the commercially available simulation tools.

The structural dynamics of the housing in turn significantly influences both the airborne sound emitted by the engine and the structure-borne sound vibration introduced into coupled structures, which at subsequent points of the propagation path can either likewise lead to airborne sound phenomena or to vibrational phenomena. For this reason too, the structural dynamics of engine components as well as bearing and housing with respect to noise behavior are of great importance. Since a consideration of this strongly coupled overall system on the simulation level has not been possible up to now, the experimental investigation and optimization of the overall system has a high priority.

The lack of acoustic feedback in relation to the load condition of the engine during operation - pull and push operation sound almost identical, only at zero load, the emitted noise noticeably reduced - is criticized as unusual. Because while the noise of an internal combustion engine has a strong dependence of the total level of load and speed, the level of an electric motor has a rather unsteady course on load and speed.

The object of the invention is therefore to solve the problems described with reference to the prior art at least partially and in particular to provide a method for operating an electric motor, with which specifically the acoustic characteristics of the electric motor can be adjusted.

This object is solved by the independent claims. The respective dependent claims are directed to advantageous developments.

In the method according to the invention for operating an electric motor in a system, for example a motor vehicle, a correction signal for changing the operating noise of the system at at least one predefinable frequency is determined from a frequency-dependent response function of the system, which is used to control the electric motor.

The inventive method is carried out in particular for certain operating points of the electric motor. The at least one predeterminable frequency is preferably the frequency of a disturbing noise or a frequency of an acoustic characteristic of the system to be changed. In particular, the desired, changed or improved noise behavior of the electric motor should be achieved by a passive method, ie by a predetermined pilot control. In contrast, or in addition to be in active processes adaptive regulations or the Feedback of an actual value used to derive the required measure for noise influence by a comparison with a nominal value (a so-called feedback control).

By means of a transfer path-specific interference suppression or noise modification, in the method according to the invention this noise behavior of the electric motor can be selectively changed without having to accept losses in terms of efficiency, weight, functionality or costs. For this purpose, the acoustic effect chain of the noise to be eliminated is preferably followed in order to be able to combine the relevant influencing variables "exciter forces", "structure-borne sound transmission property" and "radiating behavior" in conjunction with one another and resulting in a given acoustic characteristic of the electric motor during operation.

Under the consequences of the acoustic effect chain, the consideration of interactions of different effective sizes in the electric motor, which can lead to an increase or decrease of vibrations of certain frequencies, understood. Preference is given to an experimental consideration of the chain of action, wherein a simulation of the chain of action according to the invention is also possible. Under the system can here on the one hand the entire motor vehicle, on the other hand, but also the electric motor with the built-in it (stator, rotor, etc.) and / or its associated components (housing, gear, etc.) to be understood.

The consideration of the action chain will be explained below by way of example. As excitation forces here are considered primarily the forces from the magnetic flux change, which lead mainly in the radial direction, but also in tangential and sporadically in the axial direction to force fluctuations. But also other excitation forces originating, for example, from mechanical sources - such as those caused by the rotor dynamics - can be included in this way and even their interactions with the magnetic forces are taken into account. Aeroacoustic suggestions, for example due to ventilation currents, could also be considered.

In the further course of the chain of action, the transmission property and the radiation behavior are determined experimentally, above all by the structural dynamic behavior of the components, of the housing or of the housing complex involved. This means that the final "spectral fingerprint" of the noise of the electric motor in the system which is relevant for the noise characteristic of the electric motor in the system is thus considerably influenced by the components of the system and the electric motor.

This does not differ significantly from the acoustic behavior of a motor vehicle driven by an internal combustion engine. However, a fundamental difference with regard to the generation of engine noise between internal combustion engines and electric motors lies in the internal excitation and is essential for the effectiveness of the present invention. In the internal combustion engine, the ignition and incipient combustion provide the actual acoustic stimulation, which, in view of the bridging of many games, leads to a quasi-impulse excitation with an almost flat frequency spectrum (ie almost all frequencies are stimulated equally over a wide frequency range), then the relevant one Excitation of the electric motor from the force excitation of the magnetic flux and is therefore extremely tonal aligned. Accordingly, individual frequencies including appropriate higher harmonics (usually settled in the higher frequency range of a few kHz) are discretely excited.

This is where the method according to the invention comes into play, since it is possible to exert considerable influence on the noise pattern in the case of exclusively frequency-discrete ("tonal") excitation via a spectrally very fine modification (in the extreme case monofrequency change or even narrowband changes). Such noise influence by influencing a single frequency in the excitation would not be feasible in the shock excitation in the internal combustion engine, however. In the case of an electric motor, this means that because of the frequency-discrete excitation, the orders which are acoustically dominant in the noise pattern are usually not those which have the greatest amplitudes in their exciting forces. Rather, aspects such as phase angles and vibration modes of the housing play a role. It is known from experiments, for example, that a 60th engine order, which is relatively insignificant in force excitation, results in 10-15 dB higher airborne sound radiation than the 10th order and 30th order, which are dominant in relation to the excitation force. This circumstance can be explained by the fact that for this force excitation side rather insignificant 60th motor order, the phase position of the flux changes and the excited vibration mode of the engine-transmission system made this dominance possible in the first place.

This is where the inventive method begins. According to the invention, an acoustically relevant (for example, the user perceived as disturbing) order in the noise behavior of the electric motor in the system, for example in the motor vehicle or in the engine-transmission network, characterized reduce, taking into account the backward transfer path up to the electric motor feeding the effects contributing to the formation of the corresponding order considered. This inverse transfer path-specific compensation makes use of the - the corresponding noise generating - resonance effects and the hitherto often not fully understood interactions between the individual effects, because these excitation rather unobtrusive acoustic orders require minimal changes in the magnetic flux generation.

It is completely irrelevant whether the resonant overshoot in noise has its cause in the magnetic force, in the rotor dynamics or in other structural behavior, because by their interactions (eg air gap changes due to rotor vibrations, which in turn cause magnetic force changes and therefore again affect the rotor vibration) is taken into account by the compensatory influence of the control of the electric motor by the correction signal, the entire mechanism of action.

In accordance with a preferred procedure, a frequency to be changed, for example an interference frequency, is first determined here. This is done by measuring the corresponding noise and a Fourier transform, for example a fast Fourier transform, of this signal. The spectrum obtained in this way, a (mathematically) complex spectrum from an amplitude spectrum and a phase spectrum, allows the determination of the frequency to be changed, for example an interference frequency. Then a possible target noise, or its spectrum, is determined, in which, for example, the spectrum is changed or synthesized in the laboratory until the target noise is reached. Then, a measurement of the control voltage of the electric motor, for example, the frequency converter of the electric motor. Then, a frequency response function analysis is performed between the motor drive, that is, the control voltage, and the corresponding actual noise of the corresponding system. Based on this frequency response function of the system, the signal component responsible for the amplitude at the frequency to be changed can be determined. By inverting the phase of this or these signal components, a correction variable in the frequency space is obtained, from which the correction signal in the time period is obtained by a Fourier transformation. This correction signal, which was determined in advance in the laboratory, is now superimposed on the (original) control voltage of the electric motor. If the electric motor is operated with this superimposed signal, the amplitude is correspondingly reduced at the frequency to be changed, the noise is reduced or the noise of the electric motor changed accordingly. This preferred process control is carried out in particular operating point-dependent. For significant operating points, especially often occurring operating points or operating points with a particularly characteristic noise, this analysis is performed and deposited the corresponding correction signal or the superimposed signal in the control of the electric motor.

• a) Bestimmung einer Steuerspannung des Elektromotors in einem Betriebspunkt des Elektromotors; und
• b) Bestimmung der Frequency Response Funktion des Systems basierend auf der Steuerspannung.

According to an advantageous embodiment of the method, the frequency-dependent response function is determined by the following steps:

• a) determining a control voltage of the electric motor in an operating point of the electric motor; and
• b) Determining the frequency response function of the system based on the control voltage.

By means of the mathematical method of the frequency response function, it is therefore possible to determine a transmission behavior of the electric motor via or with the system (s) under consideration.

• A) Ermittlung einer Störfrequenz;
• B) Untersuchung der Responsefunktion im Hinblick auf die Signalamplitude und Phase bei der Störfrequenz;
• C) Bestimmung einer Korrekturgröße durch Phaseninvertierung; und
• D) Ermittlung des Korrektursignals durch Fourier Transformation der Korrekturgröße.

According to an advantageous embodiment of the method, the correction signal is determined by the following steps:

• A) determination of an interference frequency;
• B) examination of the response function with regard to the signal amplitude and phase at the interference frequency;
• C) determination of a correction variable by phase inversion; and
• D) Determining the correction signal by Fourier transformation of the correction quantity.

Under a disturbance frequency can also be understood in general a frequency to be changed. In this case, the response function is understood in particular to be the frequency response function of the system, in particular of the motor vehicle or the system under consideration, which represents the transmission behavior of the system. The Fourier transformation used is preferably an algorithm based on a so-called fast Fourier transformation.

According to an advantageous embodiment of the method, the correction signal is superimposed on the control voltage of the electric motor.

This is done in particular at certain operating points of the electric motor.

The method described so far represents a so-called passive method, in which a correction signal, possibly operating point-specific, is determined in advance and stored, for example, in the control of the electric motor. Alternatively or additionally, however, an active signal can also be used in which a correction signal is determined analogously to the passive method, but this during operation of the system. Here, the system can particularly advantageously represent the motor vehicle itself. This will be explained below.

According to an advantageous embodiment of the method, the at least one sound code in the system is determined at least temporarily and the correction signal is varied as a function of the deviation of the sound code from a predefinable desired code.

Here, a sound recognition of the system is recorded, in particular by microphones and / or vibration sensor. A setpoint identifier, for example a specific sound characteristic of the electric motor and / or of the motor vehicle, is stored and adjusted with the measured sound identification. The corresponding correction signal is preferably determined as described above via the frequency response function of the system.

• A) Ermittlung einer Störfrequenz;
• B) Untersuchung der Responsefunktion im Hinblick auf die Signalamplitude und Phase bei der Störfrequenz;
• C) Bestimmung einer Korrekturgröße durch Phaseninvertierung; und
• D) Ermittlung des Korrektursignals durch Fourier Transformation der Korrekturgröße.

In this context, it is advantageous if the variation of the correction signal is determined by the following steps:

• A) determination of an interference frequency;
• B) examination of the response function with regard to the signal amplitude and phase at the interference frequency;
• C) determination of a correction variable by phase inversion; and
• D) Determining the correction signal by Fourier transformation of the correction quantity.

According to a further aspect of the present invention, a method for operating a motor vehicle is proposed, which is at least temporarily driven by at least one electric motor in which determines an operating point of the electric motor and the electric motor operated at this operating point according to a method for operating an electric motor as described herein becomes.

The details and advantages described for the method of operating an electric motor can be applied and applied to the method for operating a motor vehicle.

According to a further aspect of the invention, a motor vehicle is proposed with at least one electric motor for at least temporary driving of the motor vehicle, comprising a control unit which is suitable and intended for carrying out the method for operating an electric motor as described herein.

The details and advantages described for the method for operating an electric motor and a motor vehicle can be applied and applied to the motor vehicle according to the invention and vice versa. As an electric motor in particular rotary field or traveling wave motors are used. In particular, the motor vehicle may be an electric vehicle or a hybrid vehicle operable with an internal combustion engine, in particular a gasoline engine, a diesel engine or a Wankel engine, and / or at least one electric motor.

According to an advantageous embodiment of the motor vehicle, this comprises at least one error sensor for receiving a sound recognition in the motor vehicle.

The error sensors may be microphones and / or vibration transducers. With microphones, in particular, the sound recognition can be recorded, which is based essentially on the transmission through the air, while by vibration sensor, in particular arranged in the engine compartment and / or on the body, the sound can be recorded, which is essentially on the transmission through the structures of the motor vehicle (structure-borne sound) is based. Both types of error sensors can be used independently or in combination to record the sound detection.

According to an advantageous embodiment of the motor vehicle, this comprises a frequency converter for driving the electric motor, which is connectable to the control unit.

This connection is particularly wired, but can also be done wirelessly. Preferably, the control unit controls the frequency converter according to the inventive method.

The features listed individually in the claims can be combined with one another in any technologically meaningful manner and can be supplemented by explanatory facts from the description and details from the figures, wherein further embodiments of the invention are shown.

The invention and the technical environment will be explained in more detail with reference to FIGS. The figures show particularly preferred embodiments, to which the invention is not limited. In particular, it should be noted that the figures and in particular the illustrated proportions are only schematic. Show it:

1 an example of an at least temporarily driven by an electric motor vehicle motor vehicle;

2 : Amplitude spectrum of the noise of the motor vehicle;

3 : Amplitude spectrum of a possible target noise of the motor vehicle;

4 : spectral excitation of the motor vehicle;

5 : Transfer function between the control voltage and the noise spectrum of the motor vehicle;

6 original control voltage of the electric motor;

7 original control voltage of the electric motor with correction signal;

8th Flowchart of the passive method for noise modification;

9 : Flow diagram of a passive and active method for noise modification; and

10 : an example of a motor vehicle with a combined passive and active noise modification;

1 shows a motor vehicle 1 , which by an electric motor 2 is drivable. It is a rotating field or traveling field motor. For speed control of the electric motor 2 is a frequency converter 3 formed, the available DC voltage from a battery 4 converted into an AC signal whose frequency substantially corresponds to the desired target speed. For this purpose, usually by pulse width modulation (PWM) a target speed-correlated waveform (such as a sine) is iteratively approximated, depending on the clock frequency of the frequency converter 3 more or less accurate.

The approach presented here offers a simple and very effective way of reducing the noise characteristics of systems such as motor vehicles 1 comprising an electric motor 2 to influence and in particular virtually neutral in terms of efficiency to reduce the noise of electric motors or even to suppress by this drive signal is minimally modified by superimposing compensatory correction signals.

This is made possible by the transfer path-specific inversion of the order-relevant signal components in the control, which quantitatively considered the leading to noise training overshoots in the system transmission behavior in the determination of noise suppression and thus uses them as a kind of "amplifier" in the suppression.

By an acoustic measurement of an electric motor 2 , Especially under laboratory conditions, the noise components to be reduced are determined and converted by a Fast Fourier Transformation (FFT) in an amplitude and a phase spectrum. In the relevant operating point of the electric motor 2 the time signal of the motor control (supply) is also detected and also transformed by FFT in amplitude and phase spectrum. By means of a special analysis method, for example the Frequency Response Function (FRF), the complex averaged cross-spectrum of the two signals (noise and frequency spectrum of the feed) and the separately averaged autospectrum can be used to determine the complex transfer function between the initial excitation of the feed and the noise response.

If now the noise components by the inverse transfer function - the so-called Admittanzfunktion - sent, then one obtains from this the stimulating signal contents of the motor control. If these are rotated in their phase by 180 ° and superimposed on the original motor drive, then the noise component can ideally be eliminated (canceled out by destructive superimposition).

Due to the transmission behavior of the system, which is usually very influential in the case of interfering noise, compensatory cancellation, as will be shown below, usually requires only insignificant signal changes (superpositions) in the motor supply, so that effects of efficiency do not occur.

How this transfer path-specific inversion of a noise can practically take place is illustrated by the following example: First, the level contents of the interfering / to-be-compensated orders and their phase position relative to the other signal contents are determined on the basis of an FFT of the measured noise. This is in 2 an amplitude spectrum 5 the measured noise of an electric motor 2 shown. The associated phase spectrum need not be discussed here. Clearly recognizable here are several amplitude peaks that have their highest levels in the range between about 0.5 kHz and 5 kHz. The sound of this electric motor 2 was referred to as extremely unpleasant and annoying, which is based on listening (with synthetically modified sounds) to an amplitude overshoot 6 could be moored at 3210 Hz. This amplitude increase 6 corresponds to a 48th engine order and therefore moves in their frequency with the engine speed. In 3 is an amplitude spectrum 7 a possible target noise - determined as described above by synthetic "noise design", in which the output noise is modified so long by parameterizable band stop filter until it is judged in the hearing as no longer annoying / annoying. In the present case, this had to increase the amplitude 6 this 48th order from the original amplitude peak 6 to a reduced amplitude overshoot 8th be lowered until the previously disturbing noise component in the perception was completely gone.

A measurement of the motor supply, ie the control voltage of the electric motor 2 , in the relevant operating point, it is also allowed by an FFT (see 4 ) the spectral excitation 9 of the system (motor vehicle 1 ) by the motor control. 4 shows an example in which this was determined by regenerative operation in towing operation. The responsible for the noise 48th order (see reference numeral 10 ) is discretely present in the motor drive signal but does not appear conspicuous.

In 5 the resulting transmission behavior between the motor control and the running noise of the electric motor is mapped, which can be determined by the mathematical process of the Frequency Response Function (the so-called frequency response). 5 shows the corresponding Frequency Response Function 11 of the system (of the motor vehicle 1 ), which is a gain behavior 12 the noise component of the 48th engine order. This transfer behavior involves more than just the structural dynamics of the components or assemblies involved. Rather, they are reflected in the Frequency Response Function 11 also the temporal inertia of the electrical subsystem due to capacitances and inductances, geometrical influences as well as all effects due to the deviation of the real magnetic field from the idealized field course. A determination of the overall effect can therefore only take place by determining the entire chain of effects from the initial supply of the motor to the resulting noise component. Also reciprocal approaches (ie excitation on the response side and reaction measurement on the pathogen side) can alternatively be used for this purpose. It is advantageous, however, that the entire chain of action is included in the considered transfer path and thus, by means of a suitable correlation method, the effective transfer function can be determined as a complex variable (that is, according to amount and phase).

This (mathematical) complex transfer function with amplitude and phase component between the motor supply and noise allows now to determine the necessary signal component, which is necessary to prevent the formation of the noise or reduce by the detected signal components with a phase rotation of 180 ° superimposed on the original control voltage. Thus, a correction signal can be determined, which is the control voltage of the electric motor 2 is superimposed to reduce the noise. This will be explained below, in particular with reference to the 6 and 7 be described in more detail.

In 6 is the control voltage (external conductor voltage) determined in regenerative operation during drag as described above 13 of the electric motor 2 represented, so the time signal of in 4 shown amplitude spectrum of the motor control. Even in this time domain representation, it is already possible to detect portions of higher frequencies or orders which are superimposed on the actual speed-synchronous profile.

In 7 is for direct comparison the original control voltage (the so-called outer conductor voltage signal) 13 of the electric motor 2 together with the modified signal according to the present method 14 illustrated, wherein the green curve, the original outer conductor voltage signal 13 and the red curve is the correction signal 15 reproduces. The modified signal 14 results as the superposition of the original control voltage 13 with the correction signal 15 , Where the two curves are exactly congruent creates a black curve, so that in this graph only the visible red and green curve parts indicate the extent of these necessary, minor signal changes. Accordingly, it can be assumed in such a minimal change in the motor control that this results in no measurable efficiency differences.

The advantage of this approach lies in the fact that one can relatively easily determine the overall effect of the transmission by the integral approach without losing the individual effects (parts of the structural dynamics of the components, proportions of storage and housing, proportions based on the acoustic radiation behavior of the housing, etc .) to analyze. Rather, one thus obtains the possibility of inversely determining directly the system-relevant level components responsible for the disturbance, in order to then addively superimpose these very small order components in the motor supply in such a way that a significant reduction of the disturbing noise is achieved at the end.

The deposit of the necessary for noise suppression modifications in the signal of Motor control also allows the connection of different controls. In particular, with regard to a sound design achievable with this principle, here offers the opportunity to make the acoustic feedback of the electric motor, for example, depending on the driver more clearly depending on the load. Here, instead of a noise to be suppressed, a sound to be emphasized is turned off and the procedure is analogous. A coupling of different sound patterns in a control map to existing assistance systems (eg as in the suspension control: sporty vs. neutral vs. comfortable) is also possible.

8th shows a flowchart of the heretofore described passive method for noise modification or for suppressing a noise in the noise spectrum of the electric motor 2 of a motor vehicle 1 , The procedure described is similar in its operating principle so-called active systems, but without being an active system itself. In active systems, the phenomenon to be combatted is superimposed on an effect which is corresponding in amplitude but 180 ° in its phase position, so that a so-called destructive superposition ("extinction") occurs for the phenomenon to be combated. Such active systems generally use a feedback (looped back) actual value of the sound or oscillation phenomenon to be optimized, which is compared with a defined nominal value (target value) and then the required control intervention is derived therefrom (so-called feedback control).

This comparison between an actual value to be measured and a specified setpoint, together with the intervention measure to be derived therefrom, does not take place in the previously described process control. Rather, the modification of the motor control necessary for achieving the desired target noise is already fixed once and thus statically, since the relevant parameters (for example structure-dynamic behavior of the components, radiation behavior of the housing, etc.) do not change significantly over the motor life. However, another additional active control is also possible, which will be described in detail below. This active control is also possible regardless of the previous passive noise modification.

In active control, the target noise can be actively generated or influenced. For this purpose, the above-described passive method for noise suppression or noise modification is first applied and so the necessary control voltage at the operating point quasi pre-control already held to then actively to make the sound according to the specifiable requirements ("sound design"), for example, a desired load To generate load feedback for the occupants or the driver.

In principle, the procedure is similar to passive noise removal or noise modification in which the acoustic chain of action is also considered backwards. However, the transmission behavior (or the Frequency Response Function (FRF)) is not formed as before only on the (overall) system of the electric motor, but it must be considered extended to the transfer path in a passenger compartment. Here, both airborne sound paths and structure-borne sound paths can be taken into account, whereby, in order to produce a powerfully sounding sound, above all the rather medium- and low-frequency orders and therefore, to a large extent, the structure-borne sound paths are of importance. Detailed knowledge of the necessary sound pattern (orderly mix) to associate the desired feedback characteristics (powerful, sporty, confident, comfortable, etc.) can be assumed to be known.

The necessary for generating this order mixture suggestions are "active" superimposed on the signal of the motor control. Active means in this context that on short and long-term changes in the noise spectrum in the motor vehicle 1 can be reacted by changing the correction signal. Again, to determine the necessary correction signals, the transfer path-specific transmission characteristics as described above are taken into account quantitatively in order to use the system behavior with the least possible changes in the control voltage 13 of the electric motor 2 to achieve the acoustic target.

The use of an active system is justified, above all, by a large number of possible influencing factors which can not be sufficiently flexibly taken into account by means of a passive precontrol. While, for example, different operating modes such as recuperation or boosting are registered centrally in the vehicle control unit and therefore passive pilot control can be adequately taken into account, in particular unavoidable variations in vehicle transmission behavior lead, viewed in a multitude of vehicles or in many respects variable marginal conditions for interior noise (eg vehicle occupancy). Window opening, operation of the music or hands-free system, etc.) to situations that can often not be sufficiently considered in advance. Also, any changes in the operating behavior of the electric motor or its suspension in the vehicle structure due to aging or wear can be taken into account only by an active feedback control. 9 shows a flowchart of a correspondingly combined passive and active-acting noise modification in the motor vehicle 1 ,

10 schematically shows a motor vehicle 1 with a passenger compartment 16 , The car 1 is at least temporarily via an electric motor 2 drivable. In particular, it is in the motor vehicle 1 to an electric or hybrid vehicle. In the passenger compartment 16 are first error sensor 17 educated. The first error sensor 17 are designed as microphones that control the sound in the passenger compartment 16 take up. Second error sensor 18 for example, in an engine compartment 19 of the motor vehicle 1 educated. The second error sensor 18 For example, they can be designed as microphones or vibration pickups. The error sensors 17 . 18 So serve to record a sound in the car 1 , About the first error sensors 17 In particular, the airborne sound transmission path is monitored, while the second error sensors 18 the structure-borne sound transmission path is monitored.

The signals of the first error sensors 17 and the second error sensors 18 be in a control unit 20 fed via which a frequency converter 3 of the electric motor 2 and over this the electric motor 2 self-directed. In the control unit 20 is a comparison between the actual signal, which from the error sensors 17 . 18 is recorded, and a predetermined target signal, a target noise made. In the case of deviations of the sound recognition from a predefinable setpoint identification, an analysis is carried out as described above in connection with FIG 2 to 7 portrayed a correction signal 15 supplies, which with the control voltage 13 superimposed in the frequency converter 3 is fed. This leads to an adaptation of the control voltage 13 to a modified signal 14 and thus - as described above - to a corresponding adaptation of the sound detection in the motor vehicle 1 to the predefinable target noise. In the example shown here is still a switching unit 21 formed over which the driver of the motor vehicle 1 can switch the characteristic of the target noise. The appropriate choice of the driver is then in the control unit 20 and optionally also in the frequency converter 3 fed.

The method described here allows control of an electric motor 2 in which a predefinable desired characteristic of the system can be achieved by taking into account the effective paths in the system in the sound transmission, for example via structure-borne noise by the system and / or airborne sound. In particular, the response function is the frequency response function, which represents the transmission behavior of the system. In particular, this response function is determined operating point-dependent. By this procedure, noise can be reduced and / or certain desired characteristics of the system can be impressed. The correction signal can be determined in advance once per operating point (passive method) and / or constantly or temporarily determined during operation (active method).

LIST OF REFERENCE NUMBERS

1

 motor vehicle

2

 electric motor

3

 Frequency converter

4

 battery

5

 amplitude spectrum

6

 amplitude overshoot

7

 amplitude spectrum

8th

 amplitude overshoot

9

 Spectral excitation

10

 interference

11

 Frequency response function

12

 reinforcing behavior

13

 Original control voltage

14

 Modified signal

15

 correction signal

16

 cabin

17

 First error sensor

18

 Second error sensor

19

 engine compartment

20

 control unit

21

 switching unit

Claims (10)

Method for operating an electric motor ( 2 ) in a system, for example a motor vehicle ( 1 ), in which a correction signal from a frequency-dependent response function of the system ( 15 ) is determined to change the operating noise of the system at least one predetermined frequency, which for controlling the electric motor ( 2 ) is used. Method according to Claim 1, in which the frequency-dependent response function is determined by the following steps: a) Determining a control voltage ( 13 ) of the electric motor ( 2 ) in an operating point of the electric motor ( 2 ); and b) determining the frequency response function of the system based on the control voltage ( 13 ). Method according to one of the preceding claims, in which the correction signal ( 15 ) is determined by the following steps: A) determination of an interference frequency; B) examination of the response function with regard to the signal amplitude and phase at the interference frequency; C) determination of a correction variable by phase inversion; and D) determination of the correction signal ( 15 ) by Fourier transformation of the correction quantity. Method according to one of the preceding claims, in which the correction signal ( 15 ) of the control voltage ( 13 ) of the electric motor ( 2 ) is superimposed. Method according to one of the preceding claims, in which at least one sound code in the system is determined at least temporarily and the correction signal ( 15 ) is varied as a function of the deviation of the sound detection of a predetermined desired identifier. Method according to Claim 5, in which the variation of the correction signal ( 15 ) is determined by the following steps: A) determination of an interference frequency; B) examination of the response function with regard to the signal amplitude and phase at the interference frequency; C) determination of a correction variable by phase inversion; and D) determining the correction signal ( 15 ) by Fourier transformation of the correction quantity. Method for operating a motor vehicle ( 1 ), which at least at times by at least one electric motor ( 2 ) is driven, in which an operating point of the electric motor ( 2 ) and the electric motor ( 2 ) is operated at this operating point by a method according to one of the preceding claims. Motor vehicle ( 1 ) with at least one electric motor ( 2 ) for the at least temporary drive of the motor vehicle ( 1 ), comprising a control unit ( 20 ) which is suitable and intended for carrying out the method according to one of the preceding claims. Motor vehicle ( 1 ) according to claim 8, further comprising at least one error sensor ( 17 . 18 ) for receiving a sound recognition in the motor vehicle ( 1 ). Motor vehicle ( 1 ) according to claim 8 or 9, further comprising a frequency converter ( 3 ) for driving the electric motor ( 2 ) connected to the control unit ( 20 ) is connectable.

Images