Method, auxiliary equipment and vehicle for reducing noise of three-phase current driven motor

Abstract

The invention relates to a method for reducing the noise of an electric motor driven with three-phase current, comprising the following steps: detecting the angular state of the rotor of the electric motor; calculating a reference frequency based on the detected angular state; forming sine and cosine values of the calculated reference frequency (8); filtering the formed sine and cosine values by means of a secondary path, wherein the control path to the electric motor is modeled by the secondary path by means of a transfer function and a reference signal vector is formed thereby; detecting a sound-related feedback variable on the electric motor; forming a weighting vector based on the detected feedback variable and the formed reference signal vector; forming a superposition signal from the sum of the sine and cosine values weighted by the weighting vector; and applying the superposition signal to a manipulated variable of the electric motor. A method for reducing the noise of a machine driven with three-phase current is also proposed, wherein the generation of interfering noise can be reduced even when the rotor or stator is excited.

Description

Technical Field

[0001] The invention relates to a method for reducing the noise of an electric machine driven with three-phase current, a computer-aided device for carrying out such a method, and an electrified motor vehicle. Background Art

[0002] With the increasing use of high-power electric motors, there is a need to reduce the unwanted and sometimes unpleasant noise generated by these motors, particularly in the automotive sector. Conventional sound insulation systems are largely unsuitable for this purpose, and various attempts are therefore known to reduce the impact on people, such as passengers in such motor vehicles. For example, so-called noise cancellation methods, or ANR (Active Noise Reduction), known for example from sound-isolating headphones, are known, particularly for undisturbed music enjoyment or for enabling distraction-free entertainment in small propeller-driven vehicles. To this end, there are proposals to equip the passenger compartment of vehicles with corresponding noise cancellation systems. However, these approaches do not protect against environmental influences. Effective noise reduction using these systems has proven to be very complex. Furthermore, it is difficult to address secondary sound excitations, which can also be audible and disruptive, caused by the vibration excitation of components. Summary of the Invention

[0003] Starting from this, the present invention is based on the object of at least partially overcoming the disadvantages known from the prior art. Features according to the invention are derived from a method for reducing the noise of an electric machine driven with three-phase current, a computer-aided device for implementing such a method, and an electrified motor vehicle. Advantageous embodiments are shown in preferred embodiments. Features of the invention can be combined in any technically appropriate manner, and for this purpose, features from the following description and from the drawings, including supplementary embodiments of the invention, can also be used.

[0004] The present invention relates to a method for reducing the noise of an electric motor driven with three-phase current, the method comprising the following steps:

[0005] a. From the motor, detecting the angular state of the motor's rotor;

[0006] b. Calculating a reference frequency based on the detected angle state;

[0007] c. Forming the calculated sine and cosine values of the reference frequency;

[0008] d. filtering the formed sine and cosine values by means of a secondary path, wherein the control path to the motor is modeled by means of a transfer function by the secondary path and a reference signal vector is formed therefrom;

[0009] e. detecting a sound-related feedback parameter on the motor;

[0010] f based on the detected feedback parameter and the reference signal vector formed to form a weighted vector;

[0011] g. forming a superimposed signal by the sum of the sine values and the weighted cosine values weighted by the weighting vector; and

[0012] h. Applying the superimposed signal to the manipulated variable of the electric motor.

[0013] Unless expressly stated to the contrary, ordinal numbers used in the preceding and following descriptions are for clarity only and do not reflect any order or ranking of the components referred to. An ordinal number greater than one does not necessarily imply the presence of another such component.

[0014] Here, a method for reducing the noise of an electric motor driven with three-phase current is proposed. The generation of this disturbing noise can be reduced even during the excitation of the stator and / or rotor. First, the angular state of the rotor, i.e., its relative position or angular velocity relative to the stator and the excitation coils or magnets, is detected. Subsequently, the corresponding reference frequency f is calculated. ref Then, the reference frequency f ref The product with the entire period (2π) forms the sine and cosine values. The sine and cosine values, which are preferably processed as two-dimensional vectors, are filtered in the secondary path, wherein the control path toward the motor to be regulated is modeled by the secondary path. For this purpose, for example, approximate functions or more complex simulations of the control path can be used within the scope of a digital filter. In addition, a sound-related feedback variable e is detected on the motor, which is directly related to the interfering noise. Based on the detected feedback variable e, the sine and cosine values are weighted, for example with the help of the least mean square algorithm (LMS algorithm). The superposition signal y is then generated from the sum of the weighted sine and cosine values, that is, the sine weighted value w is formed. sin and cosine weighted value w cos To this end, the formed weighted vector or its corresponding vector value (ie the sinusoidal weighted value w about the sinusoidal value) is sin and the cosine weighted value w about the cosine cos ) are respectively multiplied by the corresponding (sine or cosine) value. The sum is formed from the weighted sine product and cosine product. This corresponds, for example, to the amplitude A of the noise-canceled signal ANR With reference frequency f ref and the phase of the noise cancellation signal This superposition signal is now applied to the (multiple) manipulated variables of the motor (for example, the voltage value v according to the Park transformation of the stator). d and v q). As a result, the generation of noise that is perceived as disturbing in the motor is already significantly reduced. For example, a reduction of 16 dB (sixteen decibels) or 85% (eighty-five percent) can be achieved using this method. The above explanation is given below as a formula. The counter signal can be described mathematically using formula (1):

[0015]

[0016] Furthermore, the amplitude and phase can be determined with the aid of formulas (2) and (3):

[0017]

[0018]

[0019] In order to ensure the stability of the adaptive filter during operation, the adaptive filter is implemented as an FIR filter (finite impulse response filter, finite impulse response filter) for example. In an embodiment, the filter coefficient b of the adaptive filter is optimized in the following manner by means of an iterative optimization method in the form of an LMS algorithm, that is, the square of the superimposed signal is minimized. The signal e to be minimized is mathematically described by formula (4). Here, the filter coefficient is adapted to the increment u in the steepest descent direction. The increment affects the convergence speed on the one hand and the stability on the other hand. The filter weight value w is adjusted by means of the gradient of formula (5) sin and w cos The optimization can be performed iteratively.

[0020]

[0021]

[0022] Furthermore, in an advantageous embodiment of the method, it is provided that in step b. the reference frequency is calculated from the product of a predetermined characteristic number and the detected angular state.

[0023] In this embodiment, it is proposed to multiply the angular state of the rotor of the motor by a predetermined characteristic number to calculate the reference frequency f ref This number is related, for example, to the number of stator magnets or stator coils or to the number of phases of a motor driven with three-phase current. For example, in a three-phase motor, such an ordinal number is 6 (six), 12 (twelve), or 18 (eighteen).

[0024] Furthermore, in an advantageous embodiment of the method, it is proposed that the method comprises a plurality of parallel transmission paths, wherein steps b. to g. are respectively performed in each transmission path with a different characteristic number in step b., wherein in step g. a unique superposition signal is formed for each transmission path, and in step h. a plurality of superposition signals are applied in a superimposed manner to the manipulated variable of the stator of the electric motor.

[0025] It is proposed that the method comprises a plurality of parallel transmission paths, wherein the individual transmission paths are implemented with different ordinal numbers and are preferably implemented in the same manner as the method described above in other respects. 1a 、y 1b 、y 1c (multiple) manipulated variables (e.g. v d 、v q ).

[0026] Furthermore, in an advantageous embodiment of the method, it is provided that the angular state is a frequency signal in an electrical or mechanical reference frame of the electric machine.

[0027] Here, the angular state is a frequency signal (e.g., f el) either in the electrical reference frame or in the mechanical reference frame of the motor. The angular state is therefore directly related to the rotational speed and the queried torque and is simultaneously a signal that can be used in the method without further preprocessing.

[0028] Furthermore, in an advantageous embodiment of the method, it is provided that the feedback variable is based on at least one of the following values:

[0029] - the surface acceleration of a component driven in rotation by the motor, preferably the stator of the motor; and

[0030] An acoustic signal, which is preferably detected in the interior of a passenger compartment of a motor vehicle driven by the electric machine.

[0031] In one embodiment, the feedback variable e is the surface acceleration of the component driven in rotation, preferably directly the surface acceleration of its stator, or alternatively the surface acceleration of the rotor or a component connected to the rotor. The surface acceleration is directly related to the mechanical excitation of the rotating system and, therefore, to the generation of disruptive noise. For example, when a sudden increase in torque is applied, an increased slip relative to the excitation occurs, resulting in oscillatory excitation of the stator, stator magnets, stator coils, or corresponding cores of the stator. The same applies to the rotor, but detection in the case of the rotor is more complex due to the rotation. This, in turn, interacts with the excitation field and can cause (audible or magnetic) mechanical vibrations on the stator side. Due to the very short control path of the method proposed here, a rapid response can be achieved even to such sudden events, for example, achieving (stable) maximum damping after only 0.5 seconds (half a second), with a significant reduction of, for example, 50% already achieved in less than half the time. Additionally or alternatively, it is proposed that an acoustic signal, detected in the interior of the passenger compartment, for example, using a microphone, be used as the feedback variable e. The advantage of this embodiment is that the sound transmission can also be taken into account and, if necessary, any disruptive noise caused by existing damping effects can be omitted. Furthermore, in a preferred embodiment, a stronger weighting is applied depending on the detected sound signal while simultaneously detecting the surface acceleration, that is, in simpler terms, depending on the sound level of the sound signal or how disruptive the corresponding frequency is perceived based on the sensitivity of the human ear.

[0032] Furthermore, in an advantageous embodiment of the method, it is proposed that a transfer function of the secondary path of the electric motor is empirically derived based on the transfer characteristic of a frequency sweep of a manipulated variable applied to the stator of the electric motor, wherein the transfer function is preferably a function that approximates the transfer characteristic derived empirically.

[0033] It is now proposed that the transfer function of the secondary path be determined empirically by applying corresponding frequency sweeps. The amplitude and phase resulting from this can be detected by frequency (frequency sweep). In an advantageous embodiment, an approximation of the transfer function, i.e., a function that minimizes deviations, is determined and used in the secondary path.

[0034] Furthermore, in an advantageous embodiment of the method, it is provided that the physical control path is modeled by a secondary path, wherein the control path comprises at least one of the following components:

[0035] -Machine regulator;

[0036] -Phase converter;

[0037] - nonlinear compensation element;

[0038] - Pulse width modulation;

[0039] -Inverter;

[0040] - the dynamic electric field of the motor; and

[0041] -Measurement value detection.

[0042] In one embodiment, the secondary path corresponds to the physical control path (although preferably also to a simplified degree). In a preferred embodiment, it is not necessary to divide the secondary path into separate components to simulate the complex, largely nonlinear transfer characteristics of the control path. Rather, the definition here only refers to which part of the control path is mapped by the secondary path, and thus, in this advantageous embodiment, the secondary path is partial, or its input and output variables are correlated with the control path.

[0043] According to another aspect, a computer-aided device is provided, which has at least a processor and a memory for executing a method according to the embodiments described above.

[0044] A computer-aided device comprises one or more processors, such as a general-purpose processor (CPU) or a microprocessor, a RISC processor, a GPU and / or a DSP. A computer-aided device comprises, for example, additional elements, such as a memory interface. Alternatively or additionally, these terms refer to devices that are capable of executing provided or included programs, preferably in a standardized programming language (such as C++, JavaScript or Python), and / or controlling and / or accessing data storage devices and / or other devices (such as input interfaces and output interfaces). The term "computer-aided device" also refers to multiple processors or multiple (sub) computers that are interconnected and / or otherwise communicatively connected and that may share one or more other resources (such as memory).

[0045] The data storage is, for example, a hard disk drive (HDD, SSD, HHD) or (non-volatile) solid-state memory, such as ROM memory or flash memory (EEPROM). The memory typically consists of multiple separate physical units or is distributed across multiple separate devices, with access to the memory via data communication (e.g., Package Data Service). The latter is a decentralized solution in which the memory and processors of multiple separate computing units are used to replace or supplement the (single, modular) central onboard computer.

[0046] According to another aspect, a computer program is provided, comprising a computer program code, wherein the computer program code is executable on at least one computer in such a way as to cause the at least one computer to perform a method according to the above-described embodiments, wherein at least one of the computers:

[0047] - integrated into an on-board computer of a motor vehicle; and / or

[0048] - configured for communication with an on-board computer of a motor vehicle.

[0049] According to another aspect, a computer program product is provided, on which a computer program code is stored, wherein the computer program code is executed on at least one computer in such a way as to cause the at least one computer to perform a method according to the embodiments described above, wherein:

[0050] At least one of these computers:

[0051] - integrated into an on-board computer of a motor vehicle; and / or

[0052] - configured for communication with an on-board computer of a motor vehicle.

[0053] The computer program product containing the computer program code is, for example, a medium such as RAM, ROM, SD card, memory card, flash card, or disk. Alternatively, the computer program product is stored on a server and can be downloaded. Once the computer program is readable by a readout unit, such as a drive and / or installation program, the computer program code and the methods contained therein can be implemented by a computer or by communicating with a plurality of computer-aided devices, such as those described above.

[0054] In one embodiment, a learning algorithm is integrated, in which a plurality of data can be taken into account and used to improve the control result. Such learning algorithms (deep learning algorithms) are already known from the field of speech recognition or speech processing and face recognition and are characterized in that they are based on a data volume that cannot be fully mastered by humans and / or are based only on partially known or completely unknown rules. In contrast to finite element algorithms, such deep learning algorithms are fundamentally trivial, but due to the complexity (in this case, especially the amount of basic data), these tasks are not achievable for humans or can only be achieved with an unreasonable expenditure of time. Known deep learning algorithms or available program libraries are, for example, Keras and Cognitive Toolkit: For example, acceleration events or terrain data in a motor vehicle (in a navigation system) and the resulting noise generation in the electric motor can be predicted by a learning algorithm.

[0055] According to another aspect, an electrified motor vehicle is provided, comprising at least the following components:

[0056] - at least one driven wheel;

[0057] - an electric motor driven by three-phase current, the electric motor being connected to the at least one drive wheel in a torque-transmitting manner so as to propel the motor vehicle forward;

[0058] - at least one electrical energy storage device for the electric machine;

[0059] - an inverter for supplying voltage to the electric machine from at least one voltage source in a phase-controlled manner; and

[0060] - at least one on-board computer having a processor and a memory,

[0061] The at least one onboard computer is configured to implement the method for reducing noise of an electric motor according to the above-described embodiment.

[0062] The motor vehicle is, for example, an electrified passenger car, such as the Porsche Taycan. The driven wheels are configured to propel the motor vehicle forward and can be supplied with a corresponding torque by means of the at least one electric motor. To measure the propulsion force, a speed sensor (often referred to as an accelerator pedal) is provided in the driver's cab, for example. The speed sensor predetermines control values for a method for controlling an electric motor driven with three-phase current, using which the acceleration (and preferably also the deceleration) of the motor vehicle can be intuitively set. In some embodiments, additional input values are provided, for example, taking into account vehicle values and / or traffic data. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The invention will be described in detail below, starting from the relevant technical background and referring to the relevant drawings showing preferred design solutions. The invention is not limited in any way by the purely schematic drawings, but it should be noted that these drawings are not accurate in size and are not suitable for defining dimensional ratios. In the drawings:

[0064] Figure 1 : shows a schematic regulation loop for controlling a motor driven with three-phase current;

[0065] Figure 2 : Schematic regulation loop showing the transfer path;

[0066] Figure 3 : shows the parallel application of three parallel delivery paths;

[0067] Figure 4 : shows the transfer function of the secondary path in the Bode diagram;

[0068] Figure 5 : A graph showing a series of measurements of the system's response;

[0069] Figure 6 : a graph showing a series of measurements on the vibration of the system; and

[0070] Figure 7 : shows a motor vehicle having an electric motor. DETAILED DESCRIPTION

[0071] exist Figures 1 to 3 shows a schematic control loop or a portion thereof for a method for controlling an electric motor 1 driven with three-phase current or for reducing its noise. The method steps are represented here as blocks or with the aid of mathematical symbols. This representation corresponds to diagrams commonly used in control technology, where the individual blocks do not necessarily need to be implemented as separate hardware components, that is, elements of a hardware-implemented control loop. Instead, the control system shown is preferably performed on a computer-aided device that is configured to execute all method steps using one or more processors 37 and one or more memories 38. The computer-aided device has corresponding interfaces for indirect and / or direct communication with measuring sensors and control variable transmitters.

[0072] The motor 1 is preferably a synchronous motor. To power the motor 1, a voltage source 44 is provided, which provides a voltage, preferably a DC voltage. This DC voltage is converted by means of an inverter 35 into a three-phase current, which is fed into the coils of the stator 86 of the motor 1, causing the rotor 3 of the motor 1 to rotate. Each phase of the drive circuit 46 of the motor 1 is connected to at least one coil. The currents in the different phases are angularly offset. In the case of a three-phase drive circuit 46, the angular offset between each coil is 120°.

[0073] exist Figure 1 The schematic control loop shown in FIG. 1 merely illustrates a possible application of the noise reduction method described herein and does not describe this method in detail. The inverter 35 is controlled by a control device 47. The control device 47 provides a control data set 48 to the inverter 35. According to the illustrated embodiment, the control data set 48 includes a voltage signal for each of the three phases of the drive circuit 46 of the electric motor 1. The voltage signals are applied to the gates of the inverter 35 to influence the current in each phase.

[0074] In this embodiment, the control device 47 performs an angle estimation. This angle estimation allows sensorless determination of the electrical angle 49 of the phase of the drive circuit 46. In the case of a synchronous motor, the rotor angle of the motor 1 is inferred from this.

[0075] The sum of the superposition signal 14 and the manipulated variable 19 (processed in the machine controller 32) and the angular state 2 is described by a combined data set 50, which (after processing in the amplitude correction block 51) is the output variable of the addition block 52. In some embodiments, a plurality of superposition signals 14, 15, 16 are obtained as a result of different numbers of one or more transmission paths 23, 24, 25 and are provided as a summed superposition signal 53 (see Figure 3 ) is superimposed on the manipulated variable 19. The combined data set 50 enters as two input variables into a control calculation block 54, which processes the data used as control data in the inverter 35 and sends it as control data, for example as a voltage value, to the inverter 35.

[0076] In addition to the method steps already described, a pilot angle 55 is calculated in a pilot angle calculation block 56. Pilot angle 55 is fed as an input variable to the nonlinear compensation element 34, which compensates for system-induced errors or smoothes out fluctuations. Further input variables flow into the nonlinear compensation element 34: a correction voltage 57, which includes the DC voltage provided by the voltage source 44 and possibly other voltage values (e.g., for calculating conductor losses), a time correction value 58, and an estimated, future angle 49.

[0077] In the illustrated embodiment, the machine controller 32 calculates a reference current vector 59 with two reference current values according to the rotor-fixed coordinate system used (d / q transformation or Park transformation). A measured current vector 60, which contains two current values according to the rotor-fixed coordinate system used, is subtracted from the reference current vector 59. The measured current vector 60, which contains two current values according to the rotor-fixed coordinate system used, is derived from data measured in the drive circuit 46. In the illustrated embodiment, the measured current vector 60 corresponds to a detection data set of current data. The data of this detection data set are output by a coordinate converter 61. The coordinate converter 61, which converts the phases u, v, and w into the reference current values d and q of the reference current vector 59, is included in the detection block 62 along with the adjacent subtraction block. Thus, in the coordinate converter 61, the current signal corresponding to the system response 63 of the drive circuit 46 is converted from the phase-specific coordinate system to the rotor-fixed coordinate system.

[0078] The resulting current vector 64, generated from the difference between the reference current vector 59 and the measured current vector 60, is converted into a reference voltage vector with a voltage reference value using a proportional-integral controller. Furthermore, further control steps, such as decoupling control, are performed. This is illustrated here by a proportional-integral (PI) control element 65 and an amplitude correction block 51.

[0079] Superposition signal 14 is added (superimposed) to processed current vector 64 (here immediately after decoupling control block 66 and immediately before amplitude correction block 51). Phase-specific voltage values for the individual phases u, v, and w are calculated from the resulting combined voltage vector of combined data set 50. Compensation voltage values 67, 68, and 69 derived from nonlinear compensation element 34 are added to each phase-specific voltage value generated as an output variable from phase converter 33, and the resulting values are converted into pulse-width modulated values for control data in pulse-width modulator 70 (PWM converter) by means of pulse-width modulation, so that these pulse-width modulated values can be processed by inverter 35.

[0080] The aforementioned addition block 52 includes a proportional-integral control element 65 (PI controller), a decoupling control block 66, and an addition symbol represented by a plus sign. The control calculation block 54 includes a phase converter 33, an addition symbol connected to the phase converter 33 in phase, a nonlinear compensation element 34, and a pulse width modulator 70 (PWM converter).

[0081] In this case, the feedback variable 12 , more precisely the surface acceleration 27 , is determined (for example solely) directly at the stator 86 of the electric machine 1 by performing a measured value acquisition 36 and is transmitted to the (first) transmission path 23 .

[0082] exist Figure 2 A schematic control loop of the (first) transmission path 23 is shown in FIG. Figure 1 Based on the detected angular state 2 of the rotor 3 (here, for example, an electrical frequency signal 26), a (first) reference frequency 4f is calculated by multiplying the selected (here, first) sequence number 20. ref Then, the (first) reference frequency 4f ref The product with the entire period (2π) forms (in the upper section) the sine value 7 and (in the lower section) the cosine value 8. The sine value 7 and the cosine value 8, which are preferably processed as a two-dimensional vector, are filtered in the secondary path 9 and generate the reference signal vector 11x f,ref, wherein a control path 10 is modeled (preferably in the form of a transfer function) from a secondary path 9 to the motor 1 to be regulated. Based on a sound-related feedback variable 12e detected on the motor 1 (e.g., surface acceleration 27 of the motor 1, surface acceleration of a component of the assembly and / or a sound signal 28 of a microphone (interior 29), the sine values 7 and cosine values 8 are weighted by means of associated values of a weighting vector 13 after processing in a processing filter 71 (e.g., analog-to-digital conversion, filtering of the observed order by an adaptive peak filter and / or anti-aliasing filter). The weighting vector 13 is derived, for example, from a reference signal vector 11x by means of a least mean square algorithm 72. f,ref Then, the superposition signal y is generated from the sum of the weighted sine values 7 and the weighted cosine values 8, that is, the sine weighted values 17w are formed. sin Sum cosine weighted value 18w cos To this end, the formed weighted vector or its corresponding vector value (ie the sine weighted value 17w about the sine value 7) is sin and the cosine weighted value 18w about cosine cos ) are respectively multiplied by the corresponding (sine or cosine) value. The sum y is formed from the weighted sine and cosine products. This corresponds, for example, to the amplitude A of the noise-cancelled signal ANR With reference frequency f ref and the phase of the noise cancellation signal The superposition signal 14 is then applied to the manipulated variable 19 (n), for example to the stator 86 of the electric machine 1 .

[0083] exist Figure 3 The parallel application of three parallel transmission paths 23, 24, 25 is shown in FIG. Figure 2 . The respective first reference frequencies 4, 5, 6 result from the angular state 2 and the respective ordinal numbers 20, 21, 22. The angular state 2 and the feedback variable 12 are, for example, identical in each of the transmission paths 23, 24, 25. Preferably, the filter is adapted to the respective ordinal numbers 20, 21, 22. As output values, the individual transmission paths 23, 24, 25 each output a superposition signal 14, 15, 16, which are superimposed on one another (here added) and thus form a sum superposition signal 53. The sum superposition signal 53 is then applied to, for example, the manipulated variable 19(n) of the stator 86 of the electric motor 1.

[0084] exist Figure 4A possible transfer function of secondary path 9 is shown in the example of a Bode diagram. A first ordinate 73 (upper diagram portion) represents the amplitude axis, here expressed in decibels. A second ordinate 74 (lower diagram portion) represents the phase axis in degrees (of 360°). An abscissa 75 represents the frequency axis in Hertz. Curve 76 (bold line) of the transfer characteristic is determined empirically, for example, by applying a frequency sweep to a specific electric machine 1. Curve 77 (thin line) of the transfer function approximates curve 76 of the transfer characteristic.

[0085] exist Figure 5 A series of measured values of the reaction of a system having a specific motor 1, for example, having Figure 1 On the (third) ordinate 78, the current intensity is plotted, here in amperes, and the surface acceleration 27 is plotted, here in m / s 2 (meters per square second). The abscissa 75 shows the time, here in seconds. In this case, the regulation for noise reduction is started and stopped alternately, wherein the regulation for noise reduction is started (for example at time zero), when the curve 79 of the q current is between about 2.5 A (two and a half amperes) and 3 A, and is stopped (for example at time 6 seconds), when the curve 79 of the q current (thin line) is close to zero. The curve 80 of the surface acceleration 27 (thick line) shows a very fast and efficient regulation behavior with high stability and reproducible noise reduction. The surface acceleration 27 is increased from about 3 m / s 2 Attenuation to a maximum of 0.6m / s 2 (On average it is better.) This results in a reduction of 16 dB or 85%. (Complete) alignment is already achieved after less than 0.5 s (half a second).

[0086] exist Figure 6 A series of measured values of vibrations of a system having a specific motor 1, for example, having Figure 1 The control device 47 shown in FIG. Figure 5 As in the example, the (third) ordinate 78 (left side) is plotted with the current intensity (in amperes) on the one hand and the surface acceleration 27 (in m / s on the other hand). 2 On the (fourth) ordinate 81 (right side), the amplitude (here in volts) and the phase of the active noise reduction (in degrees) are plotted on the one hand. The abscissa 75 shows the time (in seconds). The disturbance variable is applied as a test bench vibration. It can be seen (as in Figure 5As shown in the startup state in FIG, the q current curve 79 (lower thick line) is stably maintained below the maximum of 0.6A, and the surface acceleration 27 curve 80 (lower thin dashed line) is stably maintained at a maximum of 0.8m / s 2 Below. The value of the (first) transfer path 23 (see Figure 2 ) is also stable, i.e., the curve 82 of the ANR amplitude (the upper line in the lower line, the thin line) is around 50 V (fifty volts), and the curve 83 of the ANR phase (the uppermost line, the thick dot-dash line) is around 90° (ninety degrees). To compensate for the disturbance variable, the ANR amplitude is increased by approximately 14% (fourteen percent), and the transmission path 23 reacts dynamically and reliably within the 60 seconds considered here.

[0087] exist Figure 7 , a motor vehicle 31 is shown in a schematic top view. Optionally, an (optional electric) machine 1 is arranged in the rear region, connected to the left rear drive wheel 39 and the right rear drive wheel 40 via a transmission 84 and a differential 85 for forward propulsion of the motor vehicle 31. In the front region of the motor vehicle 31, the left front drive wheel 41 and the right front drive wheel 42 are preferably steerably arranged. These drive wheels are also (optionally additionally or alternatively) connected to a second (optional electric) machine 1 for forward propulsion. An electrical energy storage 43, preferably in the form of a traction battery, and a voltage source 44 for at least one of the electric machines 1 are included (optionally between the rear drive wheels 39, 40 and the front drive wheels 41, 42). Furthermore, an onboard computer 45 is shown, comprising a (data) memory 38 and a (data) processor 37, which preferably controls the power supply to the (here, two) electric machines 1 according to the method of the previously described embodiment. In this case, for an alternative or additional embodiment of the method, a microphone for measuring value detection 36 is provided in interior 29 of passenger compartment 30 , which microphone is communicatively connected to onboard computer 45 .

[0088] A method for reducing the noise of a machine driven by three-phase current is proposed, wherein the generation of such disturbing noise can be reduced even during the excitation of the rotor or stator.

[0089] Reference Signs List

[0090] 1 Motor 32 Machine Regulator

[0091] 2 Angle state 33 phase converter

[0092] 3 Rotor 34 Nonlinear compensation element

[0093] 4 First reference frequency 35 inverter

[0094] 5 Second reference frequency 36 Measurement value detection

[0095] 6 Third reference frequency 37 processor

[0096] 7 Sine value 38 Memory

[0097] 8 Cosine value 39 Left rear drive wheel

[0098] 9 Secondary path 40 Right rear drive wheel

[0099] 10 Control path 41 Left front drive wheel

[0100] 11 Reference signal vector 42 Right front drive wheel

[0101] 12 Feedback parameters 43 Energy storage

[0102] 13 Weighted Vector 44 Voltage Source

[0103] 14 First superimposed signal 45 On-board computer

[0104] 15 Second superimposed signal 46 Driving circuit

[0105] 16 Third superimposed signal 47 Control device

[0106] 17 Sine weighted value 48 Control data group

[0107] 18 Cosine weighted value 49 Estimated, future angle

[0108] 19 manipulated parameters 50 combined data sets

[0109] 20 First ordinal number 51 Amplitude correction frame

[0110] 21 Second ordinal number 52 Addition box

[0111] 22 Third ordinal number 53 summation superposition signal

[0112] 23 First transfer path 54 Control calculation box

[0113] 24 Second transmission path 55 Pre-control angle

[0114] 25 Third transmission path 56 Pre-control angle calculation frame

[0115] 26 Frequency signal 57 Correction voltage

[0116] 27 Surface acceleration 58 Time correction value

[0117] 28 Sound signal 59 Reference current vector

[0118] 29 Internal space 60 Measure current vector

[0119] 30 Crew compartment 61 Coordinate converter

[0120] 31 Electrified Motor Vehicles 62 Detection Box

[0121] 63 System response (current signal)

[0122] 64 Current Vector

[0123] 65 Proportional Integral (PI) Control Element

[0124] 66 Decoupling adjustment frame

[0125] 67 u compensation voltage value

[0126] 68 V compensation voltage value

[0127] 69W compensation voltage value

[0128] 70 Pulse Width Modulator (PWM Converter)

[0129] 71 Processing Filter

[0130] 72 Least Mean Square Algorithm

[0131] 73 First vertical coordinate

[0132] 74 Second vertical coordinate

[0133] 75 horizontal axis

[0134] 76 Transfer characteristic curve

[0135] 77 Transfer function curve

[0136] 78 third vertical coordinate

[0137] 79 q current curve

[0138] 80 Surface acceleration curve

[0139] 81 Fourth vertical coordinate

[0140] 82 ANR amplitude curve

[0141] 83 ANR phase curve

[0142] 84 Transmission

[0143] 85 differential

[0144] 86 stator

Claims

1. A method for reducing the noise of a motor (1) driven with three-phase current, the method comprising the following steps: a. detecting the angular state (2) of the motor rotor (3) from the motor (1); b. Calculating a reference frequency (4, 5, 6) based on the detected angular state (2); c. Forming the calculated sine value (7) and cosine value (8) of the reference frequency (4, 5, 6), wherein, The sine value and the cosine value are formed by multiplying the reference frequency by the entire period 2π; d. filtering the formed sine values (7) and cosine values (8) by means of a secondary path (9), wherein the control path (10) to the electric machine (1) is modeled by the secondary path (9) by means of a transfer function and a reference signal vector (11) is formed therefrom; e. On the motor (1), detecting a sound-related feedback parameter (12); f. forming a weighted vector (13) based on the detected feedback parameter (12) and the formed reference signal vector (11); g. forming a superimposed signal (14, 15, 16) by the sum of the sine values (7) and the weighted cosine values (8) weighted by the weighting vector (13); as well as h. Applying the superimposed signal (14, 15, 16) to the manipulated variable (19) of the electric motor (1).

2. The method according to claim 1, wherein In step b., the reference frequency (4, 5, 6) is calculated from the product of a predetermined characteristic number (20, 21, 22) and the detected angular state (2).

3. The method according to claim 2, wherein: The method comprises a plurality of parallel transmission paths (23, 24, 25), wherein steps b. to g. are respectively carried out in each transmission path (23, 24, 25) with a different characteristic number (20, 21, 22) in step b., wherein in step g. a unique superposition signal (14, 15, 16) is formed for each transmission path (23, 24, 25), and In step h., a plurality of superposition signals (14, 15, 16) are applied in superimposed fashion to a manipulated variable (19) of a stator (86) of the electric machine (1).

4. The method according to any one of claims 1 to 3, wherein The angular state (2) is a frequency signal (26) in an electrical reference frame or a mechanical reference frame of the motor (1).

5. The method according to any one of claims 1 to 3, wherein The feedback variable (12) is based on at least one of the following values: - surface acceleration (27) of a component driven in rotation by the motor (1); and - Sound signal (28).

6. The method according to any one of claims 1 to 3, wherein For the electric machine (1), a transfer function of the secondary path (9) is determined empirically based on a frequency-scanned transfer characteristic of a manipulated variable (19) applied to a stator (86) for the electric machine.

7. The method according to any one of claims 1 to 3, wherein The physical control path (10) is modeled by the secondary path (9), where The control path (10) includes at least one of the following components: - a machine regulator (32); - a phase converter (33); - a non-linear compensation element (34); - Pulse width modulation; - an inverter (35); - the dynamic electric field of the motor (1); as well as -Measurement value detection (36).

8. The method according to claim 5, wherein The acoustic signal is detected in an interior space (29) of a passenger compartment (30) of a motor vehicle (31) driven by the electric motor (1).

9. The method according to claim 6, wherein This transfer function is a function that approximates a transfer characteristic obtained empirically.

10. A computer-aided device having at least A processor (37) and a memory (38) for executing the method according to any one of claims 1 to 9.

11. An electrified motor vehicle (31) comprising at least the following components: - at least one drive wheel (39, 40, 41, 42); - an electric motor (1) driven by three-phase current, the electric motor being connected to the at least one driving wheel (39, 40, 41, 42) in a torque-transmitting manner to enable the motor vehicle (31) to move forward; - at least one electrical energy storage (43) for the electric machine (1); - an inverter (35) for supplying the electric machine (1) with voltage from at least one voltage source (44) in a phase-controlled manner; and - at least one on-board computer (45) having a processor (37) and a memory (38), in, The at least one on-board computer (45) is configured to implement the method according to any one of claims 1 to 9 for reducing the noise of the electric machine (1).

Images