DE102020214644A1 - Method and device for controlling an electrical machine - Google Patents

Abstract

The invention relates to a method for controlling an electrical machine (30), the electrical machine (30) being controlled by means of a rotary encoder-free control system, an electrical rotor position angle (11) being estimated for this purpose, the electrical rotor position angle (11) being below a predetermined The speed of the electrical machine (30) is estimated by a first observer (6-1) and when the specified speed is reached or exceeded by a second observer (6-2). The invention also relates to a device (1) for controlling an electrical machine (30), comprising a control device (2), the control device (2) being set up to regulate the electrical machine (30) by means of a rotary encoder-free control, and estimating an electrical rotor position angle (11) for this purpose, the electrical rotor position angle (11) being measured below a specified speed of the electrical machine (30) by a first observer (6-1) and when the specified speed is reached or exceeded by a second observer (6 -2) is estimated.

Description

The invention relates to a method and a device for controlling an electrical machine.

It is known to control electrical machines by means of a control system without a rotary encoder. In this case, in the controlled system, a rotor position angle is not detected by means of a sensor and the rotor position angle, on the other hand, is estimated from detected phase currents. By doing without the sensor, installation space and costs can be saved.

In the lower speed range of the electric machine, however, the estimation of the rotor position angle is currently only unsatisfactorily solved.

The invention is based on the object of creating a method and a device for controlling an electrical machine, with which a rotor position angle can be better estimated, particularly in the lower speed range of the electrical machine.

The object is achieved according to the invention by a method having the features of patent claim 1 and a device having the features of patent claim 10 . Advantageous configurations of the invention result from the dependent claims.

In particular, a method for controlling an electric machine is provided, the electric machine being controlled by means of an encoder-less control system, for which purpose an electric rotor position angle is estimated, the electric rotor position angle being below a specified speed of the electric machine by means of a first observer and at Reaching or above the predetermined speed is estimated by a second observer.

Furthermore, in particular a device for controlling an electrical machine is created, comprising a control device, wherein the control device is set up to regulate the electrical machine by means of a rotary encoder-less control, and to estimate an electrical rotor position angle, the electrical rotor position angle below a predetermined speed of the electrical machine is estimated by a first observer and when the specified speed is reached or exceeded by a second observer.

The method and the device make it possible to use different observers and methods for estimating the electrical rotor position angle for different speeds. As a result, the observer used in each case or the method can be matched to a particular speed range and a behavior of the electric machine within the speed range. An estimation method can also be adapted in order to be able to use other physical and/or electrical effects, for example, in order to estimate the electrical rotor position angle. Overall, the estimation of the electrical rotor position angle can be improved as a result, in particular in a lower speed range.

The method and the device for controlling the electric machine control in particular a torque of the electric machine.

In particular, it is provided that a current rotational speed of the electrical machine is determined on the basis of phase currents detected by means of current sensors in phases that feed the electrical machine. The speed determined is compared with the specified speed, for example by means of a control device or computing device of the device. Depending on a comparison result, the first observer or the second observer is then activated by the control device or the computing device to estimate the rotor position angle.

An observer is in particular a control engineering observer (state observer) who estimates a state of the electrical machine based on a model of the electrical machine. The estimated state of the electrical machine is compared to an actual state of the electrical machine and, if there are discrepancies, at least one parameter of the model is adjusted until the estimated state and the actual state match again. A parameter adapted here is in particular the electrical rotor position angle or a variable from which the electrical rotor position angle can be derived.

In addition to the control device, the device can have a modulation control of a drive converter, which is used, for example, to control power semiconductors by means of pulse duration modulated voltages.

Parts of the device, in particular the control device, can be designed individually or combined as a combination of hardware and software, for example as program code that runs on a microcontroller or microprocessor. However, it can also be provided that parts are designed individually or combined as an application-specific integrated circuit (ASIC).

In one embodiment, it is provided that the observers are provided by means of at least one field-programmable gate array, the field-programmable gate array being configured or being configured to provide the first observer below a specified speed and to provide the second observer when the specified speed is reached or above is or will be configured. Hardware resources can be saved as a result. In addition, a field-programmable gate array enables high-frequency signal acquisition and signal processing. In this case, high-frequency should mean in particular that an associated frequency range is above a modulation frequency of a pulse duration modulation of a drive converter. In particular, high frequency refers to a frequency that is higher by a factor of 5, preferably by a factor of 10, particularly preferably by a factor of 50, compared to the modulation frequency. If a modulation frequency is 20 kHz, for example, then a frequency of the signal detection and signal processing of the field-programmable gate array is 1 MHz, for example. Since the time for reconfiguring the field-programmable gate field is in the range of microseconds to milliseconds, the reconfiguration has no effect on the electric machine, so that the change between the first observer and the second observer is not perceived by users of the electric machine when controlling. In particular, the control device includes the at least one field-programmable gate array.

Due to the parallel signal processing of a field-programmable gate array, this can advantageously process high-frequency (~MHz) sampled quantities. Due to the high-frequency sampling, sampling or aliasing effects can be reduced as disturbance variables in the signal. Any noise components of the sensors can be digitally filtered in the field-programmable gate array, so that a very high signal-to-noise ratio can be provided for further digital processing. In comparison, a microcontroller (e.g. DSP) is significantly less suitable for processing high-frequency signals, since the recorded values of the DSP have to be transported and processed by the internal analog-to-digital converter (ADC) to the processor core, which leads to propagation delay-related and preprocessing-related delays in evaluation. With a field-programmable gate array, the sensed values of the ADC are instead available to the logic units immediately, that is, in real time. Inexpensive ADCs that can generate a high-frequency digital data stream (bit stream), e.g. delta-sigma converters, can also be evaluated with field-programmable gate arrays without further measures. Overall, the electrical rotor position angle can be better estimated as a result.

In particular, the field-programmable gate array is a run-time partially (i.e., partially) reconfigurable field-programmable gate array. The reconfiguration takes place in particular by way of partial reconfiguration, in which only part of the gate array is reconfigured or reprogrammed at runtime, but other parts of the gate array retain their respective configuration or programming. The reconfiguration is carried out in particular by means of a configuration controller, which carries out and monitors the reconfiguration. The configuration control can also be provided by means of the control device or a separate control device of the device. Methods for automatic code generation can be used to reconfigure the at least one field-programmable gate field, such as the “Xilinx System Generator for DSP”, which can automatically carry out both source code modeling and a functional analysis. For example, the "System Generator for DSP" enables the modeling of FPGA functions in a Matlab Simulink environment so that the properties can be tested against other Simulink function blocks. The automatic code generation ensures that the field-programmable gate array behaves identically to Simulink modeling.

A configuration of a field-programmable gate array includes in particular instructions with which the field-programmable gate array can be programmed such that by means of the programmed (or configured or reconfigured) part of the field-programmable gate array a specific Functionality for data processing can be provided. In the present case, such a functionality includes the speed-dependent provision of the first observer and the second observer.

In one embodiment, it is provided that the electrical rotor position angle is estimated by means of the first observer via a dependency of the inductance matrix of the electrical machine on the electrical rotor position angle, with a high-frequency voltage signal being impressed on at least one phase feeding the electrical machine and a resulting high-frequency current response being recorded and is evaluated. As a result, the electrical rotor position angle can be better estimated in the lower speed range below the specified speed, since otherwise commonly used methods, in which in particular a voltage induction in the stator is evaluated, only work inadequately there. The high-frequency voltage signal is broadband, for example, so that a broadband current response is also generated.

mit ${\displaystyle \sum L}=\frac{L_d+L_q}{2}\text{und}\Delta L=\frac{L_d-L_q}{2}\text{und}{L}_{\alpha \beta }=\left({\displaystyle \sum L}\left(\begin{array}{cc}1& 0\\ 0& 1\end{array}\right)+\Delta L\left(\begin{array}{cc}cos\left(2{\gamma }_{el}\right)& sin\left(2{\gamma }_{el}\right)\\ sin\left(2{\gamma }_{el}\right)& -cos\left(2{\gamma }_{el}\right)\end{array}\right)\right)$

If the electrical machine is designed as a permanent magnet synchronous machine, for example, the following model of the electrical machine results in ß coordinates: $$\left(\begin{array}{c}{\mathrm{and}}_a\\ {\mathrm{and}}_{\beta }\end{array}\right)=R\left(\begin{array}{c}{i}_a\\ i_{\beta }\end{array}\right)+\frac{\mathrm{i.e}}{\mathrm{i.e}t}\left(\left({\displaystyle \sum L}\left(\begin{array}{cc}1& 0\\ 0& 1\end{array}\right)+\Delta L\left(\begin{array}{cc}\text{cos}\left(2g_{el}\right)& \text{sin}\left(2g_{el}\right)\\ \text{sin}\left(2g_{el}\right)& -\text{cos}\left(2g_{el}\right)\end{array}\right)\right)\left(\begin{array}{c}{i}_a\\ i_{\beta }\end{array}\right)\right)+\frac{\mathrm{i.e}}{\mathrm{i.e}t}{\mathrm{\psi }}_{PM}\left(\begin{array}{c}\text{cos}\left(2g_{el}\right)\\ \text{sin}\left(2g_{el}\right)\end{array}\right)$$

With ${\displaystyle \sum L}=\frac{L_{\mathrm{i.e}}+{\mathrm{<figure-callout\; id="2"\; label="L\; q"\; filenames="DE102020214644A1\_0005.png"\; state="\{\{state\}\}">L</figure-callout>}}_q}{2}\text{and}\Delta L=\frac{L_{\mathrm{i.e}}-{\mathrm{<figure-callout\; id="2"\; label="L\; q"\; filenames="DE102020214644A1\_0005.png"\; state="\{\{state\}\}">L</figure-callout>}}_q}{2}\text{and}{L}_{a\beta }=\left({\displaystyle \sum L}\left(\begin{array}{cc}1& 0\\ 0& 1\end{array}\right)+\Delta L\left(\begin{array}{cc}cOs\left(2g_{el}\right)& sin\left(2g_{el}\right)\\ sin\left(2g_{el}\right)& -cOs\left(2g_{el}\right)\end{array}\right)\right)$

In this case, u _α and u _β denote the voltages and i _α and i _β the currents in the electrical machine in coordinates fixed to the stator. L _d and L _q designate the inductances of the electrical machine in coordinates fixed to the rotor and γ _el designates the electrical rotor position angle. Ψ _PM designates the flux linkage of the electrical machine.

wobei $L_{\widehat{d}\widehat{q}}=\left(\left(\begin{array}{cc}{L}_d& 0\\ 0& L_q\end{array}\right)+\mathrm{\Delta }L\left(\begin{array}{cc}0& 2{\gamma }_{er}\\ 2{\gamma }_{er}& 0\end{array}\right)\right)$

After a transformation into estimated d̂q̂ coordinates, which assumes a small error γ _er = γ _el - γ̂ _el , the result with a speed ω _el is: $$\begin{array}{l}\left(\begin{array}{c}{\mathrm{and}}_{\widehat{\mathrm{i.e}}}\\ {\mathrm{and}}_{\widehat{q}}\end{array}\right)=R\left(\begin{array}{c}{i}_{\widehat{\mathrm{i.e}}}\\ i_{\widehat{q}}\end{array}\right)+{\widehat{\omega }}_{el}\left(\left(\begin{array}{cc}0& -L_q\\ -L_{\mathrm{i.e}}& 0\end{array}\right)+\mathrm{\Delta }L\left(\begin{array}{cc}-2g_{e\mathrm{right}}& 0\\ 0& 2g_{e\mathrm{right}}\end{array}\right)\right)\left(\begin{array}{c}{i}_{\widehat{\mathrm{i.e}}}\\ i_{\widehat{q}}\end{array}\right)\\ +\frac{\mathrm{i.e}}{\mathrm{i.e}t}\left(\left(\left(\begin{array}{cc}{L}_{\mathrm{i.e}}& 0\\ 0& L_q\end{array}\right)+\mathrm{\Delta }L\left(\begin{array}{cc}0& 2g_{e\mathrm{right}}\\ 2g_{e\mathrm{right}}& 0\end{array}\right)\right)\left(\begin{array}{c}{i}_{\widehat{\mathrm{i.e}}}\\ i_{\widehat{q}}\end{array}\right)\right)+{\omega }_{el}{\mathrm{\psi }}_{PM}\left(\begin{array}{c}-g_{e\mathrm{right}}\\ 1\end{array}\right)\end{array}$$

whereby $L_{\widehat{\mathrm{i.e}}\widehat{q}}=\left(\left(\begin{array}{cc}{L}_{\mathrm{i.e}}& 0\\ 0& L_q\end{array}\right)+\mathrm{\Delta }L\left(\begin{array}{cc}0& 2g_{e\mathrm{right}}\\ 2g_{e\mathrm{right}}& 0\end{array}\right)\right)$

The rotor position angle can then be estimated via γ _er via the dependence of the inductance matrix L _d̂q ̂, on γ _er , in that the first observer is used to evaluate in particular currents and current gradients that are contained in the current response caused by the high-frequency voltage signal. By comparing an ideal current response determined using the above equation with the detected current response, the value for γ _er and, via this, the electrical rotor position angle can be estimated. In particular, the model of the electrical machine is then adjusted until the ideal current response corresponds to the actually recorded current response. γ _er in particular is then adapted as a model parameter and can be read out after the ideal current response has been successfully matched to the detected current response. The adaptation can take place, for example, by means of the Newton method or by means of the gradient descent method in a manner known per se. In other words, in particular, a prediction error of a current increase expressed by γ _er is adjusted by the observer.

In one embodiment it is provided that the electrical rotor position angle is estimated by evaluating a voltage induction in the stator of the electrical machine by means of the second observer. As a result, the rotor position angle can be reliably determined in a speed range that begins with the specified speed and extends above it. In this speed range, the last term in the above-mentioned equations of the model in particular dominates, so that the electrical rotor position angle can be reliably determined here. The rotor position angle estimation takes place here also by evaluating the second observer's prediction errors in comparison to a measured current response. In order to estimate the aforementioned last term of the equation, the formation of the gradient is adjusted with respect to the evaluation of the gradient descent method.

In one embodiment it is provided that at least one of the observers is provided by means of a machine learning method or comprises at least one machine learning method. For example, the machine learning method can be a trained artificial neural network. In principle, however, other methods can also be used.

One embodiment provides that at least one speed-dependent correction term is used when estimating the electrical rotor position angle. In this way, parasitic effects can also be taken into account. This correction term includes, for example, harmonics related to the inductance matrix of the machine.

In one embodiment, it is provided that at least one additional observer is used for at least one additional speed range to estimate the electrical rotor position angle. As a result, an operating range of the electric machine can be divided into further speed ranges, which are evaluated with observers that are optimized for a speed range.

In one embodiment it is provided that the electrical machine is a permanent magnet synchronous machine.

In one embodiment it is provided that the electrical machine is a three-phase asynchronous machine. In comparison to the synchronous machine, a voltage differential equation for the rotor circuit is also considered and taken into account for the asynchronous machine.

Further features for the configuration of the device result from the description of configurations of the method. The advantages of the device are in each case the same as in the embodiments of the method.

In particular, a vehicle is also created, comprising at least one device according to one of the described embodiments. The vehicle is in particular a motor vehicle. In principle, however, the vehicle can also be another land, rail, water, air or space vehicle. An aircraft is, for example, a drone or an air taxi.

• 1 eine schematische Darstellung einer Ausführungsform der Vorrichtung zum Regeln einer elektrischen Maschine;
• 2a eine schematische Darstellung eines erfassten Strangstroms zur Verdeutlichung des Verfahrens;
• 2b eine schematische Darstellung einer zugehörigen Strangspannung zur Verdeutlichung des Verfahrens;
• 2c eine schematische Darstellung eines Trägersignals für eine Modulationsperiode zur Verdeutlichung des Verfahrens.

The invention is explained in more detail below on the basis of preferred exemplary embodiments with reference to the figures. Here show:

• 1 a schematic representation of an embodiment of the device for controlling an electrical machine;
• 2a a schematic representation of a detected phase current to clarify the method;
• 2 B a schematic representation of an associated phase voltage to clarify the method;
• 2c a schematic representation of a carrier signal for a modulation period to clarify the method.

In 1 a schematic representation of an embodiment of the device 1 for controlling an electrical machine 30 is shown. The electrical machine 30 is fed by a pulse-controlled inverter 20, for example by a two-level inverter, using a pulse duration modulation method. The method described in this disclosure is explained below using the device 1 .

The device 1 comprises a control device 2. The control device 2 uses a current sensor system 15 formed from current sensors 16u, 16v, 16w and an analog-to-digital converter 17 on phases U, V, W of the electrical machine 30 to detect phase currents lu, Iv, Iw fed. The control device 2 provides a controlled system for controlling a torque of the electrical machine 30 in a manner known per se. To regulate, the controlled system requires a current value for an electrical rotor position angle 11 of the electrical machine 30.

The control device 2 includes, in particular, a computing device 3 and a field-programmable gate array 4. The field-programmable gate array 4 has a function block 5-1 that provides modulation control. The modulation control specifies time-dependent values for phase voltages of the electrical machine 30, which are converted by the pulse-controlled inverter 20 to feed the electrical machine 30.

In a function block 5-2, a rotational speed 10 of the electric machine 30 is estimated based on the detected phase currents Iu, Iv, Iw. This is done, for example, using filters and/or a frequency analysis. The estimated speed 10 is transferred to the computing device 3, which compares the estimated speed 10 with a predetermined speed.

If the estimated speed 10 is below the specified speed, an electrical rotor position angle 11 is estimated in the function block 5-2 by means of a first observer 6-1. This applies in particular to a speed range from zero to the value of the specified speed.

If, on the other hand, the estimated speed 10 is equal to or greater than the specified speed, then the electrical rotor position angle 11 is estimated in the function block 5-2 by means of a second observer 6-2.

The value of the specified speed can be determined, for example, by means of empirical tests in which, for example, a rotor position angle is additionally detected by means of a sensor system under laboratory conditions. By comparing the detected rotor position angle and the estimated rotor position angle 11, it can be determined from which speed 10 you should switch from the first observer 6-1 to the second observer 6-2, ie from which speed the first observer 6-1 or the second observer 6-2 is better suited for estimating the electrical rotor position angle.

The estimated rotor position angle 11 is then used to regulate the torque of the electric machine 30 by means of the controlled system in a manner known per se. In particular, the estimated rotor position angle 11 is an input parameter for the modulation control, which in function block 5 - 1 determines setpoint values for the phase voltages, which are then generated by means of the pulse-controlled inverter 20 .

In particular, it is provided that the observers 6-1, 6-2 are provided by means of the field-programmable gate array 4, with the field-programmable gate array 4 being configured or being configured to provide the first observer 6-1 below a specified rotational speed and upon reaching or above the predetermined speed is configured or will be configured to provide the second observer 6-2. This is done in particular by means of a function block 5-3, which provides a configuration control in which the configurations of the two observers 6-1, 6-2 or possibly other observers for other speed ranges are stored and used to configure the field-programmable gate array 4 if necessary. In particular, the field-programmable gate array 4 is a partially reconfigurable field-programmable gate array 4 in which only a part is reconfigured and a remaining configuration can be retained. Hardware and computing resources can be saved by reconfiguring to the respectively required observer 6-1, 6-2.

Provision is made in particular for the electrical rotor position angle 11 to be estimated by means of the first observer 6-1 via a dependency of the inductance matrix of the electrical machine 30 on the electrical rotor position angle 11, with a high-frequency voltage signal 12 being impressed on at least one phase feeding the electrical machine 30 and a resulting high-frequency current response 13 is recorded and evaluated. The high-frequency voltage signal is generated, for example, by means of the modulation control in function block 5-1 and is superimposed on the pulse duration modulation. In principle, however, other methods can also be used in this speed range.

Provision is made in particular for the electrical rotor position angle 11 to be estimated by evaluating a voltage induction in the stator of the electrical machine 30 by means of the second observer 6 - 2 .

Alternatively or additionally, it can be provided that at least one of the observers 6-1, 6-2 is provided by means of a machine learning method or comprises at least one machine learning method. Such a machine learning method can be provided, for example, by means of an artificial neural network or can include one. In particular, this is a trained artificial neural network that has been trained to provide the first observer 6-1 or the second observer 6-2. Both of the observers 6-1, 6-2 can also be trained by means of an artificial Neural network are provided. The functionalities of the artificial neural networks are provided in particular by means of the field-programmable gate array 4 .

Provision can be made for at least one speed-dependent correction term 14 to be used when estimating the electrical rotor position angle 11 . The correction term 14 can be taken into account both when estimating using the first observer 6-1 and when estimating using the second observer 6-2.

Provision can be made for at least one additional observer 6-3 to be used for at least one additional speed range for estimating the electrical rotor position angle 11 . For example, a further speed range can be provided between the first observer 6-1 and the second observer 6-2, in which the electrical rotor position angle 11 is estimated by means of a further observer 6-3 set up for this purpose.

Alternatively it can be provided that the electrical machine 30 is a three-phase asynchronous machine. The basic procedure here is analogous.

In the 2a , 2 B and 2c Schematic representations are shown to clarify the process.

The phase current i is recorded at high frequency, i.e. with a sampling frequency that is greater than a modulation frequency ( 2c ) of the pulse-controlled inverter 20 ( 1 ) with which the electrical machine 30 is fed. In particular, compared to a duration of a pulse duration modulation period T, oversampling takes place by means of the respective current sensor.

In the 2a shows a current i recorded by means of a high-resolution current sensor on the strand over time t, with individual sampled values being determined by means of a linear regression during the zero-voltage states ( 2 B , u = 0) to reconstructed currents i_r and during the active voltage states ( 2 B , u > 0) are averaged to reconstructed current rises (d/dt) i_r. The mean values for the reconstructed current i_r and the reconstructed current increases (d/dt) i_r are used by the observers to estimate the rotor position angle, ie predictions for the current increases are calculated on the basis of the voltage differential equations described. A comparison of the predictions with the reconstructed increases based on the oversampling of the high-resolution current sensor results in prediction errors, which are evaluated for the estimation of the rotor position.

In the 2 B shows a voltage u' effectively dropping on the strand over time t', which corresponds to the current i. For the active stress states (u > 0), mean stresses can be calculated, which can be used when estimating the electrical rotor position angle using the observers.

Reference List

1

contraption

2

control device

3

computing device

4

field-programmable gate array

5-1

function block

5-2

function block

5-3

function block

6-1

first observer

6-2

second observer

10

rotation speed

11

electrical rotor position angle

12

high-frequency voltage signal

13

high frequency current response

14

correction term

4pm

current sensor

16v

current sensor

16w

current sensor

17

Analog to digital converter

20

pulse inverter

30

electric machine

u

strand

V

strand

W

strand

lu

string current

IV

string current

Iw

string current

i

string current

i_r

reconstructed string current

(g/dt) i_r

reconstructed mean slope of the phase current

u'

effective string tension

t

time

T

modulation period

Claims (10)

Method for controlling an electrical machine (30), the electrical machine (30) being controlled by means of a control system without a rotary encoder, an electrical rotor position angle (11) being estimated for this purpose, wherein the electrical rotor position angle (11) is estimated below a specified speed of the electrical machine (30) by a first observer (6-1) and when it reaches or above the specified speed by a second observer (6-2). procedure after claim 1 , characterized in that the observers (6-1,6-2) are provided by means of at least one field-programmable gate array (4), the field-programmable gate array (4) being configured below a predetermined speed for the provision of the first observer (6-1). is or is configured and is configured or will be configured to provide the second observer (6-2) when the predetermined speed is reached or above. procedure after claim 1 or 2 , characterized in that the electrical rotor position angle (11) is estimated by means of the first observer (6-1) via a dependency of the inductance matrix of the electrical machine (30) on the electrical rotor position angle (11), with at least one electrical machine (30 ) a high-frequency voltage signal (12) is impressed on the feeding strand and a resulting high-frequency current response (13) is recorded and evaluated. Method according to one of the preceding claims, characterized in that the electrical rotor position angle (11) is estimated by evaluating a voltage induction in the stator of the electrical machine (30) by means of the second observer (6-2). Method according to one of the preceding claims, characterized in that at least one of the observers (6-1, 6-2) is provided by means of a machine learning method or comprises at least one machine learning method. Method according to one of the preceding claims, characterized in that at least one speed-dependent correction term (14) is used when estimating the electrical rotor position angle (11). Method according to one of the preceding claims, characterized in that at least one further observer (6-3) is used for at least one further speed range for estimating the electrical rotor position angle (11). Method according to one of the preceding claims, characterized in that the electrical machine (30) is a permanent magnet synchronous machine. Procedure according to one of Claims 1 until 7 , characterized in that the electrical machine (30) is a three-phase asynchronous machine. Device (1) for controlling an electrical machine (30), comprising a control device (2), wherein the control device (2) is set up to regulate the electrical machine (30) by means of a rotary encoder-free control and to estimate an electrical rotor position angle (11) for this purpose, the electrical rotor position angle (11) being below a predetermined speed of the electrical machine ( 30) by means of a first observer (6-1) and when reaching or above the specified speed by means of a second observer (6-2).

Images