DE102017221609A1 - Compensation of non-linearities in the case of electrical current supply to an electric machine by means of an inverter - Google Patents

Abstract

The invention relates to a compensation of nonlinearities in the case of electrical current supply to an electric machine (1) by means of an inverter (6). For this purpose, a signal _injection with an oscillating signal (V _inj ) takes place during the electrical energization, and the electric _motor (1) outputs a system response (S) in the form of an electric current to the signal _injection , and values (I _abc ) are used for the determined electrical current to compensate for the non-linearities, and compensation voltages (.DELTA.U) are determined by means of the values (I _abc ), which are then impressed in the electric current then the electric motor (1).

Description

The invention relates to a method for compensation of non-linearities in an electrical current supply of an electric motor by means of an inverter. It also relates to a control device which is designed for the corresponding actuation of an inverter and a drive system for a motor vehicle with such a control device.

The compensation of non-linearities in the electrical energization of an electric motor by means of an inverter may be important, for example, to determine machine parameters or sensorless control of the electric motor or to acoustically improve the running behavior of the electric motor.

Such nonlinearities are caused by dead times in the operation of the power switches of the inverter. This includes on the one hand deliberately used dead times in the operation of the circuit breaker, for example, to prevent short circuits in the individual half-bridges of the inverter and to compensate for output capacitances and variable resistors of the circuit breaker. On the other hand, this also includes dead times, which are inherent in the design of the circuit breakers. Such dead times constitute the time delay between a first time a request for actuation of a circuit breaker in the inverter is issued (eg, "open" or "close") and a second time at which the request was actually made.

• - eine Ermittlung von Nulldurchgängen der Phasenströme, und
• - eine Ermittlung von Kompensationsspannungen.

Solutions for compensating such nonlinearities are already known. Most of these solutions include two basic functions:

• a determination of zero crossings of the phase currents, and
• - A determination of compensation voltages.

For example, solutions known as "square-wave compensation" are known. Here, the compensation voltage in each phase of the electric motor is a square wave signal. This square wave depends on the phase current in each phase. The respective value of the square wave signal can be determined by means of mathematical equations or look-up tables, so-called LUTs. The main difficulty lies in the correct determination of the zero crossings.

For example, known as "Trapeziodal Compensation" solutions are known. In this case, the compensation voltage has a trapezoidal shape in each phase of the electric motor. The respective value for the compensation voltage can be determined by means of mathematical equations or LUTs or from a combination of both. An advantage of this is that no exact knowledge about the zero crossings is required here.

For example, solutions known as "LUT-based compensation" are known. These approaches are based on the fact that the light linear characteristic is known and stored for each power switch, for each phase and for specific voltage ranges of the inverter, and that these are directly interpolated with determined (measured, calculated) values for the currents to obtain the compensation voltage ,

In addition, solutions are known which are based on spectral analyzes or on the so-called "axis-switching".

There are now a variety of solutions known to compensate for such nonlinearities. Compensation is difficult, however, if at the same time an injection of high-frequency signals in the electric motor takes place.

The object of the present invention is to improve the state of the art.

This object is achieved by the features specified in the main claims. Preferred embodiments thereof are the dependent claims.

Accordingly, a method and a control device for compensating nonlinearities in the case of electrical energization of an electric motor by means of an inverter are proposed. The electrical power supply takes place in particular in the context of a field-oriented control or control of the electric motor.

In the proposed method is carried out as part of the electrical energization of the electric motor signal injection with a vibrating (electrical) signal. This is in particular an electrical voltage with a certain (suitably high) frequency and / or (suitable) phase position. The signal can be formed, for example, rectangular (square wave signal) or sinusoidal (sinusoidal signal). The frequency can be a high frequency. The electric motor outputs a system response in the form of an electric current, in particular a phase current (= electric current through at least one of the phases of the electric motor), to this signal injection. Now, (predicated) values for the electric current, ie the system response, for the compensation of the nonlinearities are determined. This is done in particular by interpolation of values with an observer be determined. Compensation voltages are then determined by means of the (predicated) values, which are then impressed on the electric current supply of the electric motor. This compensates for the nonlinearities.

As explained in the introduction, such non-linearities are caused by dead times in the operation of the power switches of the inverter. This includes the deliberately used dead times in the operation of the circuit breaker, as well as the design-related to the circuit breakers adhering dead times. In other words, with the invention, the dead time of the power switch of the inverter is compensated.

In such an inverter is a known device for converting a direct current into an alternating current for electrical energization of the electric motor. Such an inverter can also be referred to as a DC-AC converter. Such an inverter may, for example, each per phase of the electric motor comprise two power switches, which together form a half-bridge circuit. One of the two power switches forms a so-called high-side switch and the other of the two power switches forms a so-called low-side switch. Such power scaler may be embodied, for example, as IGBTs or MOSFETs.

Accordingly, the electric motor is, in particular, an induction machine, in particular a synchronous machine. Such a synchronous machine is, for example, a permanent magnet synchronous machine (PMSM = permanent magnet synchronous motor / machine) or a synchronous reluctance machine (RSM = Reluctance Synchronous Motor / Machine).

The (predicated) values for the electric current are preferably determined by means of an observer. Such an observer is designed in particular as a mathematical or control-technical observer, ie as a corresponding mathematical system of equations. The ascertained values for the electric current (the system response) preferably consist of a basic component and a high-frequency component. These are determined in each case. Thus, the basic component is determined and the high-frequency component is determined. The (predicated) values for the electric current are then formed by forming the sum of the associated fundamental component and the associated high-frequency component. This allows a simple determination of the respective values in the control unit for the inverter or the electric motor.

Preferably, the compensation voltages are determined by means of a lookup table from the determined (predicted) values for the electric current. As a further input variable of the look-up table, in addition to the determined (predicted) values for the electric current, a (current) supply voltage of the inverter can be provided. Values may be provided as input variables for the look-up table, which values are present in an abc coordinate system of the phases a, b, c of the electric motor. This also allows a simple determination of the respective values in the control unit for the inverter or the electric motor.

The proposed controller is used to operate the inverter. It is designed to carry out the proposed method. In particular, it therefore has corresponding inputs and outputs and calculation means to perform or at least initiate the steps of the proposed method.

The proposed method and control device is used in particular in a motor vehicle. In particular, the electric motor is used there as a traction drive, ie for propulsion of the motor vehicle, or serves as an actuator of the motor vehicle, for example for actuating a steering of the motor vehicle. Therefore, it is also proposed a drive system for a motor vehicle, with the electric motor and the inverter for electrical energization of the electric motor and with the control unit for actuating the inverter.

The invention may also refer to a computer program product with instructions executing the proposed method when the computer program product is executed on a computing unit, such as a computer or microcontroller.

• 1 ein System zur elektrischen Bestromung einer E-Maschine mit einem Wechselrichter,
• 2 ein Mittel zur Ermittlung von Kompensationsspannungen.

In the following the invention is explained in detail and with reference to figures, from which further preferred embodiments and further preferred features of the invention can be removed. Here are shown in a schematic representation:

• 1 a system for electrical energization of an electric machine with an inverter,
• 2 a means for determining compensation voltages.

The system according to 1 serves as an example for the electrical energization of a synchronous machine 1 , An identical or at least comparable system can also be used for another type of electric motor, in particular for an asynchronous machine.

According to 1 we give the system a reference torque T _ref for the synchronous machine 1 specified. This can be done by an optional speed control 2 ' be determined. The reference torque T _ref represents a torque to be provided by the synchronous machine, that is, a desired torque of the synchronous machine 1 ,

An investigation 2 determined from the reference torque T _ref electrical nominal currents I _ref for the synchronous machine 1 that must be applied to the reference torque T _ref provide. These are in particular in one with a rotor / rotor flux of the synchronous machine 1 moving dq coordinate system before (then I _{dq ref} called). This from the investigation 2 output desired currents I _ref can each with an actual current I _meas be charged to realize a current control. Otherwise, there is a control.

In a controller 3 , In particular a PI controller, corresponding setpoint voltages are determined from this. In particular also in the dq coordinate system. The nominal voltages become a signal in the form of an injection voltage V _inj superimposed. Thus, a signal injection in the form of this voltage injection is realized. The injection voltage V _inj has a known frequency f _inj on. This voltage injection may be in the direction of the d-axis of the dq coordinate system of the synchronous machine 1 or in the direction of another spatial axis. The injection voltage V _inj can be rectangular (square wave) or sinusoidal (sinusoidal).

wobei θ_k der Winkel des Vektors der Injektionsspannung V_inj zu der d-Achse ist.The injection voltage V _inj can thus be expressed as follows: $${\text{V}}_{\text{inj}}={\text{V}}_{\text{inj}}\text{cos}{\mathrm{\theta }}_{\text{k}}+{\text{jv}}_{\text{inj}}\text{sin}{\mathrm{\theta }}_{\text{k}}.$$

where θ _{k is} the angle of the vector of the injection voltage V _inj to the d-axis.

The voltages thus obtained (nominal voltages with impressed injection voltage V _inj ) will be in another investigation 4 in a stator fixed α-β coordinate system of the synchronous machine 1 converted. This will turn into another means of investigation 5 for each of the phases a, b, c of the synchronous machine 1 corresponding PWM values (PWM = pulse width modulation / pulse width modulation) for actuating the associated circuit breaker of the inverter 6 determined. This provides the inverter 6 So the synchronous machine 1 with electrical currents for (intended) operation of the synchronous machine 1 , The synchronous machine 1 is done by means of the inverter 6 So electrically energized. In through the phases a, b, c of the synchronous machine 1 flowing electrical currents is also a system response, so also the answer of the synchronous machine 1 , on the injection voltage V _inj contain. This is also referred to here as current response S. This current response S can thus the actual currents in the phases a, b, c of the synchronous machine 1 be removed. Accordingly, sensor means may be provided in at least some of the phases a, b, c to determine these actual currents. The thus determined actual currents become I _meas called.

In an investigation 7 can in the coordinate system of the phases a, b, c of the synchronous machine 1 (abc coordinate system) present actual currents I _meas be converted into the α-β coordinate system. The actual currents I _meas can in an investigative 8th then be converted further into the dq coordinate system.

The actual currents I _meas become an investigative vehicle 9 fed to power reconstruction. In it are the sizes A ₁ . A ₂ . α ₁ . α ₂ the power response S determined as explained below.

The power response S is a sum of a basic component S _fundamental , also called "fundamental current", and a high-frequency component S _hf , also called "high frequency response". The current response S can thus be expressed as follows: $$\text{S}={\text{S}}_{\text{fundamental}}+{\text{S}}_{\text{hf}},$$

wobei gilt:

a₁, a₂ =

Phase des Hochfrequenzanteils S_hf (reelle Zahl),

A =

Amplitude des Hochfrequenzanteils ${\text{S}}_{\text{hf}}=\sqrt{{\text{a}}_1^2+{\text{a}}_2^2},$

f_inj =

Frequenz der Injektionsspannung V_inj ,

α =

Phasenverzug = $\text{atan}\left(\frac{{\text{a}}_2}{{\text{a}}_1}\right)$

t =

Zeit.

In this case, the high-frequency component S _hf expressed as follows: $${\text{S}}_{\text{hf}}=\text{A cos}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}+\mathrm{\alpha }\right)={\text{a}}_{\text{1}}\text{cos}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right)-{\text{a}}_2\text{sin}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right).$$

where:

a ₁ , a ₂ =

Phase of the high frequency component S _hf (real number),

A =

Amplitude of the high frequency component ${\text{S}}_{\text{hf}}=\sqrt{{\text{a}}_1^2+{\text{a}}_2^2}.$

f _inj =

Frequency of the injection voltage V _inj .

α =

Phase delay $\text{atan}\left(\frac{{\text{a}}_2}{{\text{a}}_1}\right)$

t =

Time.

$$\text{H}=\left[\begin{array}{cc}\text{cos}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right)& -\text{sin}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right)\end{array}\right].$$

From this, the vectors or matrices X and H can be formed, wherein: $$\text{X}=\left[\begin{array}{l}{\text{a}}_1\\ {\text{a}}_2\end{array}\right].$$

$$\text{H}=\left[\begin{array}{cc}\text{cos}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right)& -\text{sin}\left(2\mathrm{\pi }{\text{f}}_{\text{inj}}\text{t}\right)\end{array}\right],$$

The calculations in a control unit are performed in a time-discrete manner, that is, in several consecutive iteration steps. In the following, the index n indicates a size of the current iteration step and the index n-1 a size of the immediately preceding iteration step.

$${\text{Y}}_{\text{n}}={\text{H}}_{\text{n}}{\text{X}}_{\text{n}}+{\text{v}}_{\text{n}},$$

wobei gilt:

Y_n =

(gemessener) Wert des Hochfrequenzanteils S_hf zum Iterationsschritt n,

v_n =

Messrauschen.

For the stationary operating state of the synchronous machine 1 thus results for the iteration step n: $${\text{X}}_{\text{n}}={\text{X}}_{\text{n}-\text{1}}.$$

$${\text{Y}}_{\text{n}}={\text{H}}_{\text{n}}{\text{X}}_{\text{n}}+{\text{v}}_{\text{n}}.$$

where:

Y _n =

(measured) value of the high frequency component S _hf to the iteration step n,

v _n =

Measurement noise.

To determine the vector or the matrix H, a look-up table (= LUT) is preferably present. The instantaneous value of the vector or of the matrix H is therefore determined for each iteration step with the aid of such a LUT. This saves computing time.

wobei gilt: $${\text{K}}_{\text{n}}={\text{P}}_{\text{n}-\text{1}}{\text{H}}_{\text{n}}^{\text{T}}{\text{S}}_{\text{n}}^{-\text{1}},$$

und $${\text{S}}_{\text{n}}={\text{H}}_{\text{n}}{\text{P}}_{\text{n}-\text{1}}{\text{H}}_{\text{n}}^{\text{T}}+{\text{R}}_{\text{n}},$$

und $${\text{P}}_{\text{n}}=\left(\text{I}-{\text{K}}_{\text{n}}{\text{H}}_{\text{n}}\right){\text{P}}_{\text{n}-\text{1}}.$$

The vector or matrix X is determined each iteration step by means of the following equations forming an observer: $${\text{X}}_{\text{n}}={\text{X}}_{\text{n}-\text{1}}+{\text{K}}_{\text{n}}\left({\text{Y}}_{\text{n}}-{\text{H}}_{\text{n}}{\text{X}}_{\text{n}-\text{1}}\right).$$

where: $${\text{K}}_{\text{n}}={\text{P}}_{\text{n}-\text{1}}{\text{H}}_{\text{n}}^{\text{T}}{\text{S}}_{\text{n}}^{-\text{1}}.$$

and $${\text{S}}_{\text{n}}={\text{H}}_{\text{n}}{\text{P}}_{\text{n}-\text{1}}{\text{H}}_{\text{n}}^{\text{T}}+{\text{R}}_{\text{n}}.$$

and $${\text{P}}_{\text{n}}=\left(\text{I}-{\text{K}}_{\text{n}}{\text{H}}_{\text{n}}\right){\text{P}}_{\text{n}-\text{1}},$$

R =

eine Kovarianz des Rauschens (dies beschreibt eines Messunsicherheit und wird bevorzugt bedarfsweise eingestellt),

I =

eine Identitätsmatrix.

Here are:

R =

a covariance of the noise (this describes a measurement uncertainty and is preferably set as required),

I =

an identity matrix.

This observer, which is basically an Extended Kalman Filter, will be on investigative 9 used for signal reconstruction. Hence the amplitude A (with the values A ₁ in the d-axis and the value A ₂ in the q-axis) and the phase information α (with the values α ₁ in the d-axis and the value α ₂ in the q-axis) of the high frequency component S _hf determined.

• - magnetische Flussverkettung Ψ_pm (auf Grund der sich darin gegebenenfalls befindlichen Permanentmagnete), und
• - elektrischer Widerstand R_S des Stators, und
• - Induktivität L_d in d-Richtung, und
• - Induktivität L_q in q-Richtung.

By means of these values A ₁ . A ₂ . α ₁ . α ₂ Optionally, it is possible to have a full rank system of equations for the synchronous machine 1 and thus the following machine parameters of the synchronous machine 1 to investigate:

• - magnetic flux linkage Ψ _pm (due to the permanent magnets therein), and
• - electrical resistance R _S of the stator, and
• - inductance L _d in the d-direction, and
• - inductance L _q in q direction.

To the compensation voltages .DELTA.U To obtain compensation for the non-linearities, a predicted value I _{hf future} must first be used for the high-frequency component S _hf the current response S are determined. This predicted value I _{hf future} takes into account the dead time T _d the circuit breaker leading to the nonlinearities. In other words, it forms the value of the high frequency component S _hf that is expected to deadlock T _d lies in the future.

wobei H_n+Nd dem interpolierten Wert der Matrix H zum zukünftigen Zeitpunkt t + Td (mit T_d = Totzeit) entspricht.The predicted value I _{hf future} for the high frequency component S _hf is preferably determined by the following equation (for each observer, in d and in q direction): $${\text{I}}_{{\text{hf}}_{\text{future}}}={\text{H}}_{\text{n}+{\text{N}}_{\text{d}}}{\text{X}}_{\text{n}}.$$

where H _{n + Nd} corresponds to the interpolated value of the matrix H at the future time t + Td (with T _d = dead time).

wobei gilt:

θ =

eine derzeitige Phasenlage,

T(θ) =

eine Umrechnungsmatrix, mit: $$\text{T}\left(\mathrm{\theta }\right)=\left[\begin{array}{cc}\text{cos}\left(\mathrm{\theta }\right)& -\text{sin}\left(\mathrm{\theta }\right)\\ \text{sin}\left(\mathrm{\theta }\right)& \text{cos}\left(\mathrm{\theta }\right)\end{array}\right].$$

The phase value of the predicted value I _{hf future} in the α-β coordinate system can be determined by the following equation: $${\text{I}}_{{\text{hf}}_{{\text{future}}_{\mathrm{\alpha \beta }}}}=\text{T}\left(\mathrm{\theta }\right){\text{I}}_{{\text{hf}}_{{\text{future}}_{\text{dq}}}}.$$

where:

θ =

a current phase,

T (θ) =

a conversion matrix, with: $$\text{T}\left(\mathrm{\theta }\right)=\left[\begin{array}{cc}\text{cos}\left(\mathrm{\theta }\right)& -\text{sin}\left(\mathrm{\theta }\right)\\ \text{sin}\left(\mathrm{\theta }\right)& \text{cos}\left(\mathrm{\theta }\right)\end{array}\right],$$

ω =

die elektrische Drehgeschwindigkeit des Rotors der Synchronmaschine 1 (Rotorkreisfrequenz),

θ =

die derzeitige Phasenlage,

ωT_d =

die Verschiebung der Phasenlage um die Totzeit.

Optionally, a compensation of the phase angle θ with the rotational speed of the rotor of the synchronous machine 1 possible. Then, instead of the current phase position θ, the compensated phase position (θ + ωT _d ) is used, where:

ω =

the electrical rotational speed of the rotor of the synchronous machine 1 (Rotor angular frequency),

θ =

the current phase position,

ωT _d =

the shift of the phase position by the dead time.

The value of the basic component S _{fundamental of} the current response S can also be predicted. This predicted value is referred to below as I _{fund future} . For its prediction, reference is made to the aforementioned desired currents I _{dq ref} . This can be done using the following equation, which gives the predicted value in the α-β coordinate system: $${\text{I}}_{{\text{discovery}}_{{\text{future}}_{\mathrm{\alpha \beta }}}}=\text{T}\left(\mathrm{\theta \; +\; \omega }{\text{T}}_{\text{d}}\right){\text{I}}_{{\text{dq}}_{\text{ref}}},$$

• - Durch lineare Approximation der gemessenen Werte der Ist-Ströme I_meas für die Zukunft mit I_future = I_meas + ωT_d. Dabei wird angenommen, dass man stets in einem annähernd linearen Teil einer Sinus-Funktion arbeitet.
• - Die Soll-Ströme i_{dq ref} können mit einem Tiefpassfilter gefiltert werden, bevor sie in das α-β-Koordinatensystem transformiert werden. Dadurch kann die Dynamik des Systems besser berücksichtigt werden.

There are other possibilities to determine the predicted values I _{hf future} and I _{fund future} :

• By linear approximation of the measured values of the actual currents I _meas for the future with I _future = I _meas + ωT _d . It is assumed that one always works in an approximately linear part of a sine function.
• The desired currents i _{dq ref} can be filtered with a low-pass filter before they are transformed into the α-β coordinate system. This allows the dynamics of the system to be better taken into account.

The predicted values I _{hf future} and I _{fund future} are summed up: $${\text{I}}_{\text{ABC}}={\text{I}}_{{\text{discovery}}_{{\text{future}}_{\text{ABC}}}}+{\text{I}}_{{\text{hf}}_{{\text{future}}_{\text{ABC}}}},$$

Before or after, the predicted values I _{hf future} and I _{fund future are transformed} into the abc coordinate system. At least one conversion matrix is also used for this purpose. Thus, finally, predicted values become I _abc for the individual axes of the abc coordinate system, ie for the individual phases a, b, c of the synchronous machine 1 ,

The predicted values I _abc are used as input values of another LUT (see 2 ). This LUT contains the compensation voltages .DELTA.U in the abc coordinate system as a function of these predicted values I _abc and a supply voltage U _dc of the inverter 6 , also referred to as "DC-link voltage". This is done in the in 2 Investigation means shown 12 for determining the compensation voltages .DELTA.U containing the associated LUT. The investigative agent 12 gives per axis / phase a, b, c of the synchronous machine 1 a value ΔU _a . ΔU _b . ΔU _c as the respective compensation voltage .DELTA.U out.

This compensation voltages .DELTA.U are modulated in the system, the calculated target voltages, for example, analogous to the injection voltage V _inj , Suitable possibilities or points for modulating the compensation voltages .DELTA.U are in or on the transmission of the desired voltages from the investigator 3 to the investigation 4 or from the investigation 4 to the investigation 5 or from the investigation 5 to the inverter 6 , Depending on which option or position is used, the compensation voltages .DELTA.U converted into the respective present coordinate system by at least one corresponding conversion matrix.

The above-mentioned observer (investigative means 9 ), ie the associated equation system, can optionally also be applied to other frequencies in the actual currents of the phases a, b, c of the synchronous machine 1 to investigate. For example, so can the harmonic oscillation 3 , Grades at a trained as square wave injection voltage V _inj be determined. For this it is only necessary to adapt the vectors or matrices X and H accordingly.

The measured actual currents I _meas in the phases a, b, c of the synchronous machine 1 can for current control, as already stated above, with the desired currents I _ref from the investigation 2 will be charged. In the case of current regulation, the influence of the high-frequency component S _hf the current response S to keep low, that is in the investigative 9 determined high frequency component S _hf preferably eliminated from the determined actual currents I _meas (reference numeral 10 ). Advantage of this is that then phase distortions or bandwidth limitations in the feedback part of the current control can be minimized. The thus corrected actual currents I _meas can be determined in a determination means 11 If necessary, converted into the dq coordinate system when the desired currents I _ref also present in the dq coordinate system.

• - Die Implementierung der Ermittlung der prädizierten Werte I_{hf future} und I_{fund future} in ein Steuergerät ist in der Praxis einfach.
• - Es wird die Kompensation von Nichtlinearitäten des Wechselrichters 6 sogar beim Vorliegen von sehr hochfrequenten Signalen ermöglicht.
• - Es handelt sich um eine LUT-basierte Vorgehensweise, die die Nichtlinearitäten des Wechselrichters 6 im gesamten Strombereich des Wechselrichters 6 kompensiert.

Special advantages of the described procedure are:

• The implementation of the determination of the predicted values I _{hf future} and I _{fund future} in a control unit is simple in practice.
• - It will compensate for nonlinearities of the inverter 6 even in the presence of very high frequency signals.
• - It is a LUT-based approach, which is the nonlinearities of the inverter 6 in the entire current range of the inverter 6 compensated.

It is basically possible that instead of the investigative means shown 12 Other means or strategies for determining the compensation voltages .DELTA.U are used. Thus, for example, a compensation can be based on rectangular signals or signals in trapezoidal shape.

LIST OF REFERENCE NUMBERS

1

Electric machine, synchronous machine

2

determining means

2 '

Speed control

3

regulator

4

determining means

5

determining means

6

inverter

7

determining means

8th

determining means

9

determining means

10

elimination

11

determining means

12

determining means

Claims (7)

Method for compensating non-linearities in the case of electrical energization of an electric motor (1) by means of an inverter (6), wherein - in the case of electrical energization, a signal _injection with a vibrating signal (V _inj ) takes place, - the electric _motor (1) the signal injection outputs a system response (S) in the form of an electric current, - values (I _abc ), in particular predicted values, for the electric current for compensation of the non-linearities are determined, - compensation voltages (ΔU) are determined by means of the values (I _abc ) which are impressed during the electrical energization of the electric motor (1). Method according to Claim 1 in which the values (I _abc ) for the electric current are determined by means of an observer. Method according to Claim 1 or 2 in which the values (I _abc ) consist of a basic component (I _fund ) and a high-frequency component (I _hf ) of the electrical current, which are determined in each case. Method according to one of the preceding claims, wherein the compensation voltages (ΔU) are determined by means of a look-up table from the values (I _abc ) for the electric current. Control device for actuating an inverter (6), characterized in that the control device is designed for carrying out the method according to one of the preceding claims. Drive system for a motor vehicle, with an electric motor (1) and with an inverter (6) for electrical energization of the electric motor (1), characterized by a control unit according to Claim 5 , Computer program product with instructions that performs the procedure according to one of Claims 1 to 4 executes when the computer program product is executed on a computing unit.

Images