WO2024028315A1

WO2024028315A1 - Device for determining the angular position of a rotor of a rotary electric machine

Info

Publication number: WO2024028315A1
Application number: PCT/EP2023/071278
Authority: WO
Inventors: Laurent Labiste
Original assignee: Valeo Equipements Electriques Moteur
Priority date: 2022-08-02
Filing date: 2023-08-01
Publication date: 2024-02-08
Also published as: FR3138744A1

Abstract

Disclosed is a device (100) for determining the angular position of a rotor (4) of a rotary electric machine on the basis of only two sensor signals delivered by position sensors (20), this device comprising: - a circuit (103) for estimating the position of the rotor, in particular via implementation of a control loop, this circuit (103) delivering as output a signal representative of the position of the rotor, and - a circuit (102) for dynamically compensating for an orthogonality defect between a first input signal (200) obtained from one of the two sensor signals delivered by the one or more position sensors (20) and a second input signal (201) obtained from the other of the two sensor signals delivered by the one or more position sensors (20), this circuit (102) for compensating for an orthogonality defect receiving as input at least: the first input signal (200) and second input signal (201), this circuit (102) for compensating for an orthogonality defect delivering as output an input of the circuit (103) for estimating the position of the rotor, this circuit (102) for compensating for an orthogonality defect being configured to: - dynamically determine, on the basis of the first and second input signals, the value of the orthogonality defect between these two signals, - construct a compensation matrix on the basis of this value thus determined, and - compensate for the orthogonality defect between said input signals (200, 201) by applying this compensation matrix to said input signals.

Description

Device for determining the angular position of a rotating electrical machine rotor

The present invention relates to a device for determining the angular position of a rotor of a rotating electric machine, as well as an assembly comprising such a determination device and such a rotating electric machine.

The electrical machine is for example an alternator or alternator powered by a nominal voltage of 12V or 48V, or even more. The electric machine can also be a propulsion machine powered by a nominal voltage of 12V or 48V, or even more, for example a voltage greater than 300V, for example 400V or 800V.

This electric machine can be integrated into a vehicle with hybrid or purely electric propulsion, for example an automobile. More generally, “vehicle” encompasses for the purposes of this application any form of mobility with purely electric, hybrid, thermal or other propulsion. “Vehicle” thus includes a vehicle rolling on land via four, three, two wheels or any other number of wheels, or a vehicle moving in the air or on water, or even in space.

Control of this electric machine requires knowledge of the angular position of the machine's rotor. To do this, it is known, in the case of a three-phase synchronous machine, to use two signals supplied by sensors, for example Hall effect, inductive type or at the end of the shaft, signals which are offset by 90 ° (electrical or mechanical), depending on the type of sensor. The signals provided by these sensors are processed by a rotor position estimation circuit, for example via a control loop. This circuit outputs a signal representative of this position of the rotor. This measurement can then be used, for example, to control the inverter/rectifier interposed between the stator of this electrical machine and the electrical energy storage unit of the vehicle's on-board network, which is in particular a battery.

The accuracy of the position obtained at the output of the rotor position estimation circuit depends, among other things, on respecting the value of 90° for the angular offset, or phase, between the two signals provided by the position sensors. A lack of orthogonality in this offset, which is therefore different from 90°, is for example caused by imprecise positioning of the sensors on the electrical machine. However, an offset other than 90° can generate errors on the second harmonic, for example of the signal representative of the position of the rotor obtained in the end, ultimately leading to oscillations of current and torque in the electrical machine.

To avoid such problems, it is known to carry out a calibration operation at the end line, just before departure for delivery of the electrical machine to the customer. However, such a calibration operation at the end of the line is complex and time-consuming, relying for example on an iterative extraction of Fourier series carried out at constant speed which are then subject to complex processing. Furthermore and above all, such a calibration operation at the end of the line is not possible when the electrical machine is supplied by an equipment manufacturer independently of its power and control electronics.

There is thus a need to remedy the aforementioned drawbacks.

The invention achieves this, according to one of its aspects, by means of a device for determining the angular position of a rotor of a rotating electric machine on the basis of only two sensor signals supplied by sensors position, the device comprising:

- a circuit estimating the position of the rotor, in particular via the creation of a control loop, this circuit providing at output a signal representative of the position of the rotor, and

- a dynamic compensation circuit for orthogonality fault between a first input signal coming from one of the two sensor signals supplied by the position sensor(s) and a second input signal coming from the other of the two signals sensor provided by the position sensor(s), this orthogonality fault compensation circuit receiving as input at least the first and second input signals, this orthogonality fault compensation circuit providing as output a transmitted signal towards the circuit estimating the position of the rotor, in particular providing an input to the circuit estimating the position of the rotor, this orthogonality fault compensation circuit being configured to:

- determine dynamically on the basis of the first and second input signals the value of the orthogonality defect between these two signals,

- construct a compensation matrix on the basis of this orthogonality defect value thus determined, and

- compensate for the lack of orthogonality between said input signals by applying this compensation matrix to said input signals.

The invention makes it possible to compensate for the lack of orthogonality arising from the position sensors used, other than via a calibration operation at the end of the line. At the output of the orthogonality fault compensation circuit, two signals can be obtained whose fundamentals are shifted by 90°. This orthogonality fault compensation is carried out upstream of the circuit estimating the position of the rotor, so that the latter receives then at input two signals whose fundamentals are shifted by 90°. The invention is based on dynamic compensation, so that it makes it possible to detect, and even correct, any position drift resulting from the aging of the machine. Although an end-of-line calibration operation is not necessary with the invention, the invention can make it possible, when such an operation is carried out, to verify that it has been carried out correctly.

The improvement in the precision of the signal representative of the position of the rotor is obtained according to the invention using a simple solution, which is implemented in the determination device upstream of the circuit estimating the position of the rotor, in particular via a control loop. Furthermore, this solution has the following advantages:

- not to require over-sampling,

- not to generate a delay in data acquisition, and therefore not to cause a risk of instability,

- not to use a complex adaptive frequency filter,

- not to involve trigonometric functions or complex cross-correlations

- not to require imposing a constant speed on the machine for a significant period of time as required by a calibration operation at the end of the line,

- not to be a static solution, but dynamic.

The solution according to the invention can provide the advantage of not requiring a state observer, such as for example a Kalman filter.

The solution according to the invention can also provide the following advantages:

- be insensitive to electrical harmonics

- be uncorrelated from the bandwidth of the circuit estimating the position of the rotor,

The signal representative of the position of the rotor is for example an angle value relative to a reference position. This may be the angle characterizing the electrical position or the angle characterizing the mechanical position of the rotor of the electrical machine.

The orthogonality fault compensation circuit is arranged upstream of the circuit estimating the position of the rotor, in particular via the creation of a control loop. The output of the orthogonality fault compensation circuit is for example directly received at the input of the circuit estimating the position of the rotor.

The number of sensor signals may be equal to the number of position sensors, with one position sensor providing, for example, only one sensor signal. Alternatively, the number of sensor signals may be different from the number of position sensors. For example, in the case where the sensors are used in differential mode, the number of position sensors is double the number of sensor signals. Where appropriate, the angular offset between the sensor signals may be equal, to the nearest ratio of the number of pole pairs, to the angular offset between sensors from which these sensor signals originate. The angular offset between the sensor signals may be equal to the angular offset between sensors from which these sensor signals originate

The orthogonality error compensation circuit can dynamically determine the orthogonality error related to the mechanical angular offset between the two sensor signals provided by the position sensors. This is for example the case when these sensors are sensors at the end of a shaft.

Alternatively, the orthogonality compensation circuit may dynamically determine the orthogonality due to the electrical angular offset between the two sensor signals provided by the position sensors. This is for example the case when the position sensors are inductive sensors, or Hall effect sensors or resolvers.

In all of the above, the dynamic determination of the lack of orthogonality between the first and the second input signal can include:

- the multiplication of the first and the second input signal and the application of a first filter, in particular low pass, to the signal resulting from this product,

- the sum of the square of the first input signal and the square of the second input signal and the application of a second filter, in particular low pass, to the signal resulting from this sum.

The output of the first filter may be proportional to the square of the amplitude and the sine of the orthogonality defect between the first and second input signals.

Taking into account the signal at the output of the second filter can make it possible to reduce the impact of a calculation error carried out at the input of the first filter. This consideration can also make it possible to improve the response time in transient conditions.

The first filter and the second filter may be different, the first filter being for example a filter of order 3 or order 2 and the second filter being in particular a filter of order 2 or 1, at least one of the first and the second filter being in particular a Butterworth filter.

Alternatively, the first and second filter may be identical, each a filter of order 1, order 2 or order 3. Each of this first and second filter can be produced in the form of a filter of order 1, order 2 or order 3. Butterworth.

Each of the first filter and the second filter is for example a low-pass filter. As will be seen later, these filters can have a constant cutoff frequency or a variable cutoff frequency. For the same moment, or even for all instant, the cutoff frequency can be the same between the first filter and the second filter.

The output of the first filter can be divided by the output of the second filter. The signal resulting from this division can attack a saturation block, making it possible to simply improve precision during the transient regime.

In all of the above, the compensation matrix can be a 2*2 matrix, with at least two of its coefficients depending on the lack of orthogonality.

The compensation matrix has, for example, coefficients

[Math. 1] y(y

where (p corresponds to the lack of orthogonality between the first and the second input signal as determined.

In the theoretical case where the value determined for the lack of orthogonality between the first and second input signals would be zero, we see that this compensation matrix would be the identity matrix h

In all of the above, compensation for the orthogonality defect can be carried out for any speed of rotation of the rotor.

Alternatively, in all of the above, the compensation for the orthogonality defect is only carried out for a range of rotor rotation speed, the orthogonality defect compensation circuit also receiving as input the rotor rotation speed . This rotation speed of the rotor is for example obtained at the output of the circuit estimating the position of the rotor. Compensation for the lack of orthogonality is only carried out, for example, above a minimum rotor rotation speed. Below this minimum rotation speed, which is for example 100, 200, 300, 400, 500, 600 or 700 rpm, no compensation for the lack of orthogonality between the first and second signals is then carried out. , and these first and second signals are directly transmitted to the circuit estimating the position of the rotor. Not compensating for orthogonality at low speeds can help avoid having to use overly selective low-pass filters.

Whether the compensation for the orthogonality defect is carried out over the entire operating range of the machine or not, when the first filter and the second filter are each a low-pass filter, the cut-off frequency of each filter can vary depending on speed. This cutoff frequency can vary continuously or discontinuously, then successively occupying constant values. The fact of varying each cutoff frequency, or at least one of these two cutoff frequencies such as that of the first filter only or that of the second filter only, can allow to achieve a compromise between precision and speed of response of the orthogonality defect compensation circuit. For example, higher cutoff frequencies of the first and second filter for lower speeds make it possible to quickly determine the value of the orthogonality defect while lower cutoff frequencies for higher speeds improve precision.

In all of the above, the circuit estimating the position of the rotor can implement a control loop and subtract from the signal at the output of this control loop a position error signal determined on the basis of the orthogonality defect between the first and second input signals.

In all of the above, the device can also include a dynamic normalization circuit by the amplitude of the first harmonic of each sensor signal supplied by the position sensor(s), this dynamic normalization circuit providing output signals to the orthogonality fault compensation circuit, providing in particular the first and second input signals of the orthogonality fault compensation circuit. This dynamic normalization circuit is thus arranged upstream of the orthogonality fault compensation circuit. This dynamic normalization circuit is for example produced according to the teaching of application W02021/121770.

The invention also relates, according to another of its aspects, to a set comprising:

- a rotating electric machine for propelling a hybrid or electric vehicle, and

- a control device for this electrical machine, comprising a determination device as defined above.

The rotating electric machine is for example a synchronous machine, for example a three-phase synchronous machine or a synchronous machine whose electrical stator winding defines a double three-phase system. The electrical stator winding is for example formed by wires or by conductive bars connected to each other.

In all of the above, the rotor may be a claw rotor. This rotor then comprises a first and a second interlocking pole wheels, the first pole wheel defining a series of claws of generally trapezoidal shape, each claw extending axially in the direction of the second pole wheel, the second pole wheel defining a series of claws of generally trapezoidal shape, each claw extending axially towards the first pole wheel. A permanent magnet can be received between two consecutive claws circumferentially speaking for the rotor.

Alternatively, the rotor may be other than a claw rotor, for example comprising a pack of sheets or being a cage rotor. In all of the above, the rotor can comprise any number of pairs of poles, for example three, four, six or eight pairs of poles.

In all of the above, the electrical machine may include a stator cooling circuit in which fluid such as air or liquid circulates. This liquid can be water or oil.

The rotor can be cooled by this same cooling circuit or by another cooling circuit in which air, or liquid such as water or oil, circulates.

The electric machine may have a nominal mechanical power of between 4 kW and 35 kW, being for example 4 kW, 8 kW, 15 kW, 25 kW or 35 kW, or the electric machine may have a nominal mechanical power of between 40 kW and 400kW, being for example 40kW, 80 kW, 100 kW, 150 kW, 180 kW, 200 kW, 300kW or 400 kW.

This rotating electric machine can be electrically powered from an electrical energy storage unit via an inverter/rectifier of the assembly, this inverter/rectifier allowing, depending on whether the electric machine operates as a motor or generator, to charge a vehicle on-board network or to be electrically supplied from this network.

The nominal voltage of the electrical energy storage unit may be 12 V, 48 V or have another value, for example another value greater than 300 V.

The rotating electric machine may also include a pulley or any other means of connection to the rest of the powertrain of the vehicle. The electric machine is for example connected, in particular via a belt, to the crankshaft of the vehicle's thermal engine. Alternatively, the electric machine is connected to other locations of the powertrain, for example at the input of the gearbox from the point of view of the torque passing to the wheels of the vehicle, at the output of the gearbox from the point from the point of view of the torque passing towards the wheels of the vehicle, at the level of the gearbox from the point of view of the torque passing towards the wheels of the vehicle, or even on the front axle or the rear axle of this powertrain.

The rotating electric machine is not necessarily a synchronous machine, it may be an asynchronous machine.

The invention also relates, according to another of its aspects, to a method for determining the angular position of a rotor of a rotating electric machine on the basis of only two sensor signals provided by position sensors, in which we use a determination device as defined previously.

All or part of what has been mentioned previously still applies to this other aspect of the invention.

This determination process is for example integrated into a process for controlling the electric machine, in which the angular position of the rotor determined as above is used to control the motor torque and/or the current of the electrical energy storage unit.

This determination method can be implemented by a computer program product, which, when read by the electric rotating machine control device, makes it possible to implement the steps of carrying out said method, said program product d computer which can be stored on a medium readable by said control device.

The invention can be better understood on reading the description which follows, a non-limiting example of its implementation and on examining the appended drawing in which:

[Fig.l] represents schematically and in axial section an example of a rotating electrical machine to which the invention can be applied,

[Fig.2] represents in elevation another type of rotor than that of Figure 1

[Fig.3] schematically represents the device for determining the position of the rotor of the machine according to a non-limiting example of implementation of the invention, [Fig.4] represents schematically and in detail an example of implementation of the orthogonality fault compensation circuit,

[Fig.5] schematically represents a detail of Figure 4,

[Fig.6] to [Fig.8] represent examples of filters that can be used to produce the part of the compensation circuit of Figure 5,

[Fig.9] schematically represents another detail of Figure 4,

[Fig.10] schematically represents yet another detail of Figure 4, and

[Fig.11] corresponds to response curves making it possible to illustrate an effect of the invention.

Figure 1 shows a polyphase rotating electric machine 1, in particular for a motor vehicle, to which the invention can be applied.

This rotating electrical machine can form an alternator or altemo-starter of the vehicle. This rotating electric machine can be powered via an electronic component of power 9 comprising an inverter/rectifier by a battery whose nominal voltage is 12 V or 48 V or a value greater than 300 V, for example.

The rotating electric machine 1 comprises a casing 2. Inside this casing 2, it further comprises a shaft 3, a rotor 4 integral in rotation with the shaft 3 and a stator 5 surrounding the rotor 4. The rotational movement of the rotor 4 takes place around an axis X. In this example, the casing 2 comprises a front bearing 6 and a rear bearing 7 which are assembled together. These bearings 6, 7 are hollow in shape and each centrally carries a ball bearing 10, H respective for rotational mounting of shaft 3.

A pulley 12 is in the example considered fixed on a front end of the shaft 3, at the level of the front bearing 6, for example using a nut resting on the bottom of the cavity of this pulley. This pulley 12 makes it possible to transmit the rotational movement to the shaft 3 and it can be connected via a belt to the crankshaft of the vehicle's thermal engine.

The rear end of shaft 3 carries, here, slip rings belonging to a collector and connected by direct connections to the winding. Brushes belonging to a brush holder 8 are arranged so as to rub on the slip rings.

The front bearing 6 and the rear bearing 7 may also include substantially lateral openings for the passage of air in order to allow cooling of the rotating electrical machine by air circulation generated by the rotation of a fan front 13 on the front dorsal face of the rotor 4, that is to say at the level of the front bearing 6 and a rear fan 14 on the rear dorsal face of the rotor, that is to say at the level of the bearing rear 7.

In this exemplary embodiment, the stator 5 comprises a body 15 in the form of a pack of sheets provided with notches, for example of the semi-closed or open type, equipped with notch insulation for mounting the electrical winding polyphase of the stator. Each phase includes a winding 16 passing through the notches of the body 15 and forming, with all the phases, a front bun and a rear bun on either side of the stator body. The windings 16 are for example obtained from a continuous wire covered with enamel or from bar-shaped conductive elements such as pins connected together. The electrical winding of the stator is for example three-phase, then implementing a star or triangle assembly whose outputs are connected to the electronic power component 9.

Rotor 4 in Figure 1 is a claw rotor. It comprises two pole wheels 17. The first pole wheel 17 faces the electronic power component 9 while the second pole wheel 17 faces the pulley 12.

Each of the pole wheels 17 comprises a bottom 18 extending radially on either side of the axis X, the wheel defining a series of claws 19 of generally trapezoidal shape. Each claw of a pole wheel 17 extends axially towards the other pole wheel from a base arranged on the radially outer periphery of the bottom 18.

The rotor 4 also includes, between the radially interior portions 20 and the claws 19, a coil wound on a coil insulator 22.

The rotor 4 can also include permanent magnets (not shown) interposed between two neighboring claws 19 at the external periphery of the rotor. Alternatively, the rotor 4 may be devoid of such permanent magnets.

The rotor 4 can also be different from that shown in Figure 1, being for example formed by a stack of sheets, as shown in Figure 2.

The number of pairs of poles defined by the rotor 4 can be arbitrary, for example equal to three, four, six or eight.

The machine also includes sensors 20 for measuring the position of the rotor 4, for example two Hall effect sensors, grouped in the same plastic housing and arranged so as to present between them an electrical angular offset equal to 90°. One of these sensors provides a first sensor signal “S” here being a sine signal, and the other of these sensors provides a second sensor signal “C” here being a cosine signal.

These sensors are for example positioned at the rear bearing 7 of the machine and they interact with a magnetic target secured to the rotation of the rotor.

The sensor signals S and C are used by a device for controlling the rotating electric machine 1, comprising a device 100 for determining the angular position of the rotor 4. The control device for the rotating electric machine 1 can be a computer for example. example a microcontroller, or an integrated circuit for example an FPGA or an ASIC, capable of receiving electrical signals and processing them, in particular sensor signals 20.

The two signals S and C supplied by these two sensors 20 are used by the device 100 for determining the angular position of the rotor 4, which will now be described with reference to Figures 3 and following.

In known manner, the device 100 comprises a circuit 101 carrying out dynamic bias compensation (“offset” in English) by averaging the samples.

In the example of Figure 3, the output signal of this circuit 101 attacks a dynamic normalization circuit 105 by the amplitude of the first harmonic of each signal coming from a position sensor 20. This circuit 105 is for example as described in application W02021/121770. As can be seen in Figure 3, the circuit 105 also receives as input an output from a circuit 102 for dynamic orthogonality defect compensation which will be described subsequently. One of the output signals from circuit 105 and coming from the first sensor signal S constitutes a first input signal 200 for the circuit 102 for dynamic compensation of the orthogonality defect, and another signal output from circuit 105 and coming from the second sensor signal C constitutes a second input signal 201 for this circuit 102. As will be described later, the circuit 102 dynamically compensates for the lack of orthogonality between the first input signal 200 and the second input signal 201, orthogonality defect which is identical to that between the first sensor signal S and the second sensor signal C.

In an example not shown, circuit 105 is not present, so that the output signal of circuit 101 is received as is at the input of circuit 102 for dynamic orthogonality defect compensation.

As can be seen in Figure 3, the circuit 102 for dynamic orthogonality defect compensation can also, but in a non-limiting manner, receive as input the rotation speed of the rotor as estimated at the output of a circuit 103 which will be described later.

In addition to an output signal received at the input of circuit 105, there are at the output of circuit 102 for dynamic orthogonality defect compensation two signals 204 and 205 respectively coming from the first 200 and the second 201 input signal, these two signals 204 and 205 being angularly offset by 90°. In the case where the first input signal 200 has the expression

[Math. 2] S=sin(omega*t) and where the second input signal 201 has the expression [Math. 3] C=cos(omega*t+phi), the signals 204 and 205 then have the expression respectively [Math. 4] S=sin(omega*t) and [Math. 5] C=cos(omega*t).

These two signals 204 and 205 are received at the input of the circuit 103 estimating the position of the rotor, this circuit 103 providing at output a signal 206 representative of the position of the rotor 4, via an angle 0 measured with respect to a reference position of this rotor. This angle corresponds in the example described to the electrical angle characterizing the position of rotor 4, but could alternatively correspond to the mechanical angle characterizing this position. This circuit 103 here implements a control loop for the position of rotor 4.

We also find at the output of circuit 102 an error signal in position 207 determined by circuit 102 on the basis of the lack of orthogonality between its first 200 and its second 201 input signal. As can be seen in Figure 3, this error signal in position 207 is subtracted from signal 206 to provide the signal representative of the position of the rotor at the output of device 100.

Circuit 103 is for example identical to that described in application W02021/121770 already cited.

We will now describe in more detail in Figures 3 to 6 the circuit 102 for dynamic orthogonality defect compensation. As can be seen in Figure 3, this circuit 102 also receives as input, in addition to the output signals from circuit 105, or where appropriate from circuit 101 directly, an output 208 from circuit 103, namely the signal representative of the speed of rotor 4.

The operation of an example of dynamic orthogonality fault compensation circuit 102 will now be described.

As can be seen in Figure 3, this circuit 102 for dynamic orthogonality defect compensation includes:

- a block 220 making it possible to determine dynamically on the basis of the first 200 and the second 201 input signal the value of the orthogonality fault between these two input signals,

- a block 221 for constructing a compensation matrix on the basis of this orthogonality defect value thus determined, and

- a block 222 for compensating for the lack of orthogonality between said input signals 200 and 201 by applying this compensation matrix to said input signals 200 and 201.

We will describe with reference to Figure 4 an example of dynamic determination block 220. This block 220 receives as input:

- the first 200 and the second 201 input signal of the dynamic compensation circuit 102,

- a filter cutoff frequency value generated by a block 224 itself receiving as input the signal 208 representative of the rotation speed of the rotor at the output of circuit 103,

- an initialization value 209 for the orthogonality fault between the first and second signals. This initialization value can be 0° or between -10° and 10°, in particular between -1° and 1°, being for example of the order of 0.5° or -0.5°,

- an initialization value 210 for the outputs of filters 212 and 213.

This block 220 applies two distinct processing operations to the first signal 200 and the second signal 201.

According to a first processing, the first signal 200 is multiplied with the second signal 201 and the resulting product is processed by a first low-pass filter 212 whose cutoff frequency value is provided by the block 224. Below a minimum rotation speed, which is for example 100, 200, 300, 400, 500, 600 or 700 rpm, the block 224 can neutralize the block 220 so that the input signals 200 and 201 are transmitted as is at the output of circuit 102 for dynamic orthogonality defect compensation.

According to a second processing, the first signal 200 is squared, the second signal 201 is squared, and the sum of these two squares is processed by a second low-pass filter 213 whose cutoff frequency is identical to that of the first low pass filter. Each of the first filter 212 and the second filter 213 is for example a low-pass filter. The first filter 212 and the second filter 213 may be different, the first filter being for example a filter of order 3 or order 2 and the second filter being in particular a filter of order 2 or 1, at least one of the first and second filter being in particular a Butterworth filter. Alternatively, the first 212 and the second 213 filter may be identical, each a filter of order 1, order 2 or order 3. Each of this first and second example filter can be produced in the form of a Butterworth filter.

Each of the low-pass filters can have a variable cutoff frequency depending on the rotation speed of the rotor. For example, this frequency varies piecewise, taking higher values for low rotation speeds and decreasing stepwise as the rotation speed increases. For any instant, the cutoff frequency is for example the same between the first filter 212 and the second filter 213.

Alternatively, each filter 212 and 213 may have a constant cutoff frequency.

Examples of filters which can constitute the first filter 212 or the second filter 213 are shown in Figures 6 to 8. Each of the filters 212, 213 uses for example one of the solutions according to one of Figures 6 to 8.

As can be seen in Figure 5, in the example described, the output of the first filter 212 is divided by the output of the second filter 213. The signal resulting from this division attacks a saturation block 215, allowing a simple way to improve the precision of the determination of orthogonality defect by the block 220 during the transient regime. At the output of this saturation block 215, a gain 216 can then be applied, here a gain of value "-2", the value of this gain being advantageously chosen so that at the output of this gain we obtain exactly the value of the fault orthogonality between the first input signal 200 and the second input signal 201.

This block 220 has an output 217 which attacks the block 221 for constructing the orthogonality defect matrix which is represented in FIG. 9. The output 217 corresponds to the value determined by the block 220 for the orthogonality defect cp between the first signal 200 and the second signal 202. The example of Figure 9 is such that the compensation matrix at the output of block 221 is a 2*2 matrix, whose coefficients are the

As can be seen in Figure 5 and as already indicated previously, the output 217 of block 220 is received at the input of the dynamic normalization circuit 105. The output 217 of block 220 is also, in the example considered, received as input to a block 223 which is shown in Figure 10. This block 223 applies in this example a gain, the value of which is for example 1/ 2, to the value determined for the orthogonality defect, and the output of this block 223 provides the signal 207 which has already been mentioned with reference to Figure 3.

Block 222 consists of applying to the first signal 200 and the second signal 201 at the input of the dynamic compensation circuit 102 the compensation matrix generated by block 221. At the output of this block 222 we obtain the signals 204 and 205 which are angularly offset of 90° and which attack circuit 103.

Figure 11 represents on the same time scale on the abscissa on a graph 300 the evolution of the rotation speed of the rotor of the electric machine as provided by the device 100,

- on a graph 301 the evolution of the input signals of the orthogonality fault compensation circuit 102, each coming from one of the position sensors 20,

- on a graph 302 the evolution of the value determined by block 220 of circuit 102 for the lack of orthogonality between the first input signal 200 and the second input signal 201; And

- on a graph 303 the evolution of the rotor position error

In this example, the two sensor signals coming from the sensors 20 have an electrical offset not equal to 90° but with an orthogonality defect of 7°.

The dynamic orthogonality fault compensation circuit 102, an example of implementation of which has been described above, is started at a time to.

We see by observing:

- curve 302 that the value of 7° is obtained for the lack of orthogonality between the input signals in less than 100ms,

- curve 303 that the position error is canceled in the same time interval, and

- curve 300 shows that the estimated rotation speed improves after compensation for the orthogonality defect, as evidenced by the disappearance of harmonics of order 2. The invention is not limited to the example of implementation which has just been described.

In particular, the dynamic orthogonality fault compensation circuit 102 can act independently of the speed of rotation of the rotor, that is to say from the start of the electrical machine. In the latter case, the cutoff frequency of each filter 212 and 213 can remain constant. Alternatively, in the latter case, the cut-off frequency of each filter 212 and 213 can vary, as described previously.

The implementation of the determination of the angular position of the rotor 4 can be carried out by a computer program product, which, when executed by the control device of the rotating electrical machine 1, makes it possible to carry out the method of determining the angular position of the rotor 4 of the rotating electrical machine 1 on the basis of the two sensor signals 20 provided by position sensors, as described with reference to Figures 3 to 11.

This computer program product can be recorded on a storage medium readable by the control device in order to carry out the implementation of the determination of the angular position of the rotor 4.

Claims

1. Device for determining (100) the angular position of a rotor (4) of a rotating electric machine on the basis of only two sensor signals (S, C) supplied by position sensors (20), device comprising :

- a circuit (103) estimating the position of the rotor, in particular via the creation of a control loop, this circuit (103) providing at output a signal representative of the position of the rotor, and

- a circuit (102) for dynamic compensation of orthogonality error between a first input signal (200) coming from one of the two sensor signals supplied by the position sensor(s) (20) and a second signal input (201) from the other of the two sensor signals supplied by the position sensor(s) (20), this orthogonality defect compensation circuit (102) receiving as input at least: the first (200) and the second (201) input signal, this circuit (102) for compensating the orthogonality defect providing as output a signal transmitted to the circuit (103) estimating the position of the rotor, notably providing an input to the circuit (103) estimating the position of the rotor, this circuit (102) for compensating for the orthogonality defect being configured to:

- construct a compensation matrix on the basis of the value of this defect thus determined, and

- compensate for the lack of orthogonality between said input signals (200, 201) by applying this compensation matrix to said input signals.

2. Device according to claim 1, in which the dynamic compensation circuit (102) dynamically determines the lack of orthogonality caused by the mechanical angular offset between the two sensor signals provided by the position sensors (20).

3. Device according to claim 1, in which the dynamic compensation circuit (102) dynamically determines the orthogonality defect linked to the electrical angular offset between the two sensor signals supplied by the position sensors (20).

4. Device according to any one of the preceding claims, in which the dynamic determination of the lack of orthogonality between the first (200) and the second (201) input signal comprises: - the multiplication of the first (200) and the second (201) signal and the application of a first filter (212), in particular low pass, to the signal resulting from this product,

- the sum of the square of the first signal (200) and the square of the second signal (201) and the application of a second filter (213), in particular low pass, to the signal resulting from this sum.

5. Device according to claim, in which the first filter (212) and the second filter (213) are different, the first filter (212) being for example a filter of order 3 or order 2, and the second filter (213) being in particular a fiber of order 2 or 1, at least one of the first (212) and the second fiber (213) being in particular a Butterworth fiber.

6. Device according to claim 4 or 5, wherein the output of the first filter (212) is divided by the output of the second filter (213).

7. Device according to any one of the preceding claims, in which the passage matrix is a 2*2 matrix, at least two of its coefficients being a function of the lack of orthogonality (cp).

8. Device according to any one of the preceding claims, in which compensation for the lack of orthogonality is carried out for any speed of rotation of the rotor.

9. Device according to any one of claims 1 to 7, in which the compensation for the orthogonality defect is only carried out for a range of rotor rotation speed, the circuit (102) for dynamic defect compensation orthogonality also receiving as input the rotation speed of the rotor.

10. Device according to any one of claims 4 to 6 and according to claim 9, each of the first filter (212) and the second filter (213) being a low-pass filter, and the cutoff frequency of each filter (212 , 213) varying as a function of speed, notably decreasing as speed increases.

11. Device according to any one of the preceding claims, in which the circuit

(103) estimating the position of the rotor implements a control loop and subtracts from the signal at the output of this control loop a position error signal determined on the basis of the lack of orthogonality between the first and the second signal d 'entrance.

12. Device according to any one of the preceding claims, further comprising a circuit (105) for dynamic normalization by the amplitude of the first harmonic of each sensor signal coming from a position sensor (20), this circuit providing output of the signals to the dynamic orthogonality defect compensation circuit (102), in particular providing the first and second input signals to the dynamic orthogonality defect compensation circuit (102).

13. Set including:

- a rotating electric machine for propelling a hybrid or electric vehicle, and - a device for controlling this electric machine, comprising a determination device (100) according to any one of the preceding claims.

14. Method for determining the angular position of a rotor (4) of a rotating electric machine on the basis of only two sensor signals supplied by position sensors (20), in which a determination device (100) is used ) according to any one of claims 1 to 12.

15. Computer program product comprising instructions which cause the assembly according to claim 13 to execute the steps of the method according to claim 14.

16. Computer-readable medium, on which the computer program product according to claim 15 is recorded.