Automatic calibration method for an angular position sensor

Abstract

Method for automatically calibrating an angular position sensor. One aspect of the invention relates to a method for automatically calibrating an angular position sensor (2) of a magnet rotor relative to a stator of a synchronous machine, comprising an angle error compensator (6) adjusting the angles measured (sensor) by the angular position sensor (2) by an angular difference (calculated by imposing a forward setpoint current equal to zero and comparing a setpoint voltage (V') with a quadratic return current (Iq), a forward return current (Id) resulting from a direct Park transformation of the current measured at the stator phase outputs and an electrical speed of the rotor, a predetermined value corresponding to the flux of the rotor magnets, and a quadratic inductance. Figure to be published with the abbreviation: Figure 1

Description

Title of the invention: Method for automatically calibrating an angular position sensor TECHNICAL FIELD OF THE INVENTION

[0001] The technical field of the invention is that of rotating electrical machines provided with an angular position sensor making it possible to measure an angular position of a rotor relative to a stator.

[0002] The present invention relates to the initial adjustment of this sensor, for example a resolver or a Hall effect sensor. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0003] An electrical machine is equipped with an inductor stator comprising windings and a rotor comprising an induced body and a rotating shaft on which is mounted a position sensor such as a resolver or Hall effect. The angular position sensor comprises a fixed part and a movable part mounted on the rotor shaft of the electrical machine which produces at the output a set of electrical signals whose relative amplitude characteristics instantly reflect the angular position of the rotor of the machine.

[0004] In a self-controlled electrical machine, this signal is used to control the current in the windings of the inductor stator to maintain an optimum angular difference (typically in quadrature) between the rotating magnetic field in the stator and the magnetic field generated between the poles of the rotor.

[0005] When the torque is very high, it may happen that the rotor shaft slips relative to the rotor body. The initial angular setting of the angular sensor is therefore modified. In addition, the movable part of the angular sensor can also slip on the rotor shaft, causing a shift in the angular position.

[0006] This shift causes a malfunction in the control of the electrical machine. Indeed, in a synchronous machine, for example three-phase, with permanent magnets, the torque produced depends on the interaction between the rotor flux and the stator flux. Since the rotor flux is produced by the permanent magnets, the torque is adjusted by adjusting the stator flux for which two parameters are accessible: the amplitude of the flux, itself adjusted by the amplitude of the currents of the three-phase power supply system, and the phase of the stator flux relative to the rotor flux. This phase is itself adjusted by the phase of the stator currents. At a given current amplitude, the maximum torque is obtained when the rotor flux is, more or less, in phase quadrature relative to the stator flux. The amplitude of the currents is controlled using regulators which use measurements from current sensors.

[0007] To be able to precisely adjust the phase of the stator currents, it is therefore necessary to have a motor rotor position sensor to know the angular position of the rotor (therefore of the rotor flux) relative to the stator winding. This is the function of the position sensor to measure the absolute position of its moving part and if there is no slippage between the moving part of the sensor and the rotor poles, the angular position of the machine rotor. A given reference position between the machine rotor and the machine stator corresponds to a reference value, for example "zero degrees", for the measurement of the rotor position by the position sensor.For the control of the electric machine and by convention, the angular reference "zero degrees" of the absolute position of the rotor of the machine on an electric revolution occurs when the rotor poles of the machine are aligned on the respective axes of the winding of a phase A, that is to say one of the phases of the stator, of which we obviously know the spatial position by the construction of the windings.

[0008] In the case of slippage between the moving part of the position sensor and the active part of the rotor, there is a difference between the indication of the measurement of the position sensor and the reference value of the rotor position. Let us call this difference "position sensor timing error".

[0009] It is therefore necessary to know this position sensor setting error in order to know exactly the position of the machine rotor flux and therefore to be able to optimize the control of the machine.

[0010] One solution is to have a method for searching for position errors in static position consisting of injecting a constant current, for example, into two phases of the stator winding. The synchronous electric machine must be no-load, that is to say that the rotor of the electric machine must be decoupled from the body of the turbomachine's thruster allowing the free rotating rotor to take a well-defined equilibrium position, the rotor flux of the magnets naturally aligning with the stator flux produced and therefore a known predetermined angular position between the rotor and the stator. The position sensor measures and transmits the measured angular position and a calculation unit can thus compare the measured angular position with the predetermined angular position at this equilibrium position of the rotor, and deduce the angular position measurement error between this angular position measurement and the predetermined angular position.The calculation unit can thus calculate, in normal operating mode of the electric machine, an adjusted angle by offsetting the angle measured by the position sensor by this angular position error.

[0011] This method of searching for position errors in static position has the disadvantage of injecting static currents into the stator at high amplitude to allow the rotor to rotate in order to obtain the known (reference) rotor angular position, causing heating of the winding and the electronic switches. of power, important during the error search procedure which can damage several elements (electronics, electronic power switches, winding, rotor magnet) of the electrical machine. In addition, for a multipolar machine with p pairs of poles, there are p electrical revolutions per mechanical revolution, there are therefore as many positions of known angular equilibrium as there are pairs of poles by supplying the two phases. It may therefore be necessary to carry out this position error search process several times in a static position by supplying, for example, other phases or repeating the same supply of the two phases to check the angular position error. It is therefore necessary to carry out the position error search process several times in a static position, for several positions resulting in a very long process.

[0012] The objective of the invention is to correct the position sensor calibration error in a simple and faster manner and by reducing the risk of damage to the various elements of the machine (electronic switches, winding, etc.). Summary of the invention

[0013] The invention offers a solution to the problems mentioned above, by making it possible to estimate under very specific conditions, while having the rotor rotating, the angular position at an instant t of the rotor relative to the stator and to compare this estimated angular position with the angular position measured by the position sensor to identify and correct the angular offset which caused the setting error of the position sensor.

[0014] One aspect of the invention relates to a method for automatically calibrating an angular position sensor measuring an angle corresponding to the mechanical angular position of an induced rotor with a magnet forming P pairs of poles relative to a stator comprising X phases of a multi-phase synchronous electric motor of an electrical machine of an aircraft thruster, the position sensor comprising a movable part mounted on the induced rotor and a fixed part relative to the stator, the method using a regulation loop to transmit a command to an inverter converter of the electrical machine to supply the phases of the stator,the control loop comprising a current regulator comprising a setpoint input for receiving a current setpoint comprising a quadratic current comprising a component on the quadratic axis of the setpoint current vector and a direct current corresponding to a component on the direct axis of the setpoint current vector and a direct setpoint current for delivering a voltage setpoint comprising a component on the direct axis of the voltage vector and a component on the quadratic axis of the voltage vector, a direct Park and inverted Park calculation unit, making it possible to transform the voltage setpoint into control of the converter, inverter, the method being characterized in that it comprises a part of identifying the error of the position sensor by the speed, a part of which is carried out by an angle error compensator, comprising the following steps: • calculate the inverter converter control in order to rotate the induced rotor at a predetermined mechanical speed in a constant manner, • measure the phase currents and transform them in a Park frame of reference, into quadratic return current and direct return current, taking into account a calculated adjusted angle, • calculate an electrical speed of the rotor by multiplying the number p of pairs of poles by the predetermined mechanical speed, • calculate a voltage setpoint during a period t in which the rotor rotates at the predetermined mechanical speed by imposing a direct current value equal to 0 on the current regulator, • Identify a shift error of the direct component on its direct axis by comparing, during the period t, the direct voltage of the voltage setpoint to a value of inverse sign, the value being equal to the calculated electrical speed multiplied by a quadratic inductance multiplied by the quadratic return current, • Identify an offset error of the quadratic component on its quadratic axis by comparing, during the period t, the quadratic voltage of the voltage setpoint with a first value equal to the calculated electrical speed of the rotor multiplied by a predetermined value corresponding to the flux of the rotor magnets, • calculate an angular difference value from these identified direct and quadratic angular offset errors, • determine at period t, an angle adjusted according to the calculated angular difference value, and the value of the measured angle, • transmit the return quadratic current and the return direct current according to the angle adjusted to the current regulator.

[0015] Thanks to the invention, the method proposes a control law which identifies the presence of an error between the actual angular position of the rotor of the electric machine and between the angular position which is used in the current regulation loop of the electric machine in a condition where the direct setpoint current is equal to zero.

[0016] Indeed, when the direct current is equal to zero, and if the position sensor does not have any positioning error of its zero angle relative to the rotor, the direct voltage at the output of the regulator comprises a value equal to - weieeLqiq and Vg comprises a value equal to weiec <Pf, Ainsi en imposant la composante du vecteur current on the direct axis of the current noted as equal to zero, if there is a positioning measurement offset between the sensor and the rotor, the angular position has no impact on the motor torque, the motor torque is therefore controlled by the component of the current vector on the quadrature axis, noted iq. Imposing the direct current equal to 0 to the regulator therefore allows, in the event of a measurement error, to identify the positioning error and to adjust the positioning to converge the component on the direct axis of the voltage vector towards its real value equal to - WeleeLgiq and converge the component on the quadrature axis of the voltage vector V towards its real value œeiec(Pf

[0017] Furthermore, such a method operates in both directions of rotation of the electrical machine, whether in the positive or negative trigonometric direction (clockwise).

[0018] In addition to the characteristics which have just been mentioned in the preceding paragraph, the method according to one aspect of the invention may have one or more complementary characteristics among the following, considered individually or according to all technically possible combinations:

[0019] According to one embodiment, the position sensor is a resolver.

[0020] According to another embodiment, the position sensor is a Hall effect sensor.

[0021] According to one embodiment, the angular difference value (A0) is calculated according to this formula:

[0022] [Math.l] A 2 [ + ( kp + -f • with 8d = ( <oeiecLqiq + V d correspondant à la comparaison pour identifier l’erreur angulaire de la composante direct et t'g = - ^eiecVf + Vg, correspondant à la comparaison pour identifier l’erreur angulaire de la composante quadratique, Lq est l’inductance quadratique, k_p,k_i sont les gains du correcteur de type proportionnel intégrale utilisée pour annuler l’erreur entre respectivement le courant quadratique de consigne et le courant quadratique retour et le courant direct de consigne et le courant direct retour.

[0023] According to one embodiment, the rotor is coupled to a shaft of the propeller comprising an inertial torque, in which during the period t, the value of the direct voltage is imposed to be equal to 0.

[0024] According to one embodiment, the rotor is coupled to a body of a turbomachine, and in that the part of calibrating the position sensor by the speed is repeated a number X times at different predetermined speeds and in that each time that a part of calibrating the position sensor by the speed is carried out, the angular difference value between the adjusted angle and the measured angle and the corresponding predetermined speed is recorded. • According to an example of this embodiment, the first predetermined mechanical speed is the highest speed of the predetermined speeds, for example the maximum rotational speed of the rotor, and in that the method restarts the part of calibrating the position sensor by the speed by decreasing the mechanical speed of the rotor by a predetermined difference until the rotor stops rotating.

[0025] According to another embodiment than the previous one, the electric machine can rotate at no load, for example by disengaging from a turbomachine shaft, the method comprises: • in the step of calculating an inverter converter command of the error identification part of the position sensor by the speed, the predetermined mechanical rotation speed is incremented in a predetermined manner each time the method restarts the error identification part of the position sensor by the speed, • in the error identification part of the position sensor by speed, a step of recording each angular difference value between the calculated adjusted angle value and the measured angle.

[0026] According to an example of this embodiment, the method further comprises, after the part of calibrating the position sensor by speed, a part of verifying the setting of the angular position sensor by current comprising the following steps: • calculate a command for the inverter converter to supply the X phases in order to rotate the induced rotor at a second predetermined mechanical speed constantly, • measure the phase currents and transform them in a Park reference frame, into quadratic return current and direct return current, taking into account the calculated adjusted angle value, • calculate an electrical speed of the rotor by multiplying the number of pole pairs by the second predetermined mechanical speed, • impose at the predetermined mechanical speed, at the input of a current regulator, a quadratic setpoint current at a first predetermined value different from zero, and a direct setpoint current at a zero value, • determine the quadratic inductance, • compare to the second predetermined speed constantly, that at the output of the current regulator, the value of the direct voltage of reference is equal to a value of inverse sign, the value being equal to the product of the calculated electric speed, the quadratic inductance and the quadratic current of back, • compare to the second predetermined mechanical speed constantly, that at the output of the current regulator, the value of the quadratic voltage of reference is equal to the product of the calculated electrical speed times the predetermined value corresponding to the flux of the rotor magnets, • calculate the angular difference between the calculated adjusted angle and the measured angle, while the rotor rotates at the second predetermined mechanical speed constantly according to these comparisons, and record it, • wherein the automatic calibration method repeats a predetermined number of times the part of checking the calibration of the angular position sensor by the current by incrementing each time the predetermined value of the setpoint quadratic current in a predetermined manner until a predetermined maximum return quadratic current value is obtained.

[0027] According to one embodiment, the command is calculated by calculating the speed according to the position measurements by the position sensor, and the current measurements,

[0028] According to one embodiment, the method comprises a step of determining the quadratic inductance from the values ​​of the direct and quadratic components of the measured current Id and Iq and from predetermined maps of the machine.

[0029] Another aspect of the invention relates to a method for controlling an electrical machine of a turbomachine, in which at each start-up of the electrical machine, the method comprises the method for automatically setting an angular position sensor described previously with or without characteristics of the embodiments described previously, and in which the method further comprises a normal mode of control of the electric motor comprising a step of calculating the control of the inverter converter to supply the phases of the stator of the electric motor according to the received setpoint of the values ​​of quadratic return current and direct return current, the calculation unit carrying out the calculations by the mathematical method of the direct or inverse Park transform, according to an angle B comprising the measured angle added to a correction value according to a recorded angular difference.

[0030] According to one embodiment, the correction value may be equal to the recorded calculated angular difference value corresponding to the recorded predetermined mechanical speed closest to the measured mechanical instantaneous speed.

[0031] According to another embodiment, the method further comprises a step of calculating an average of the recorded angular difference values ​​and in which the correction value is equal to the average of the recorded angular difference values.

[0032] Another aspect of the invention relates to a control loop of an electric machine of a turbomachine for calculating an inverter converter command. for an electric motor comprising an inductor stator having stator windings forming X phases, an induced rotor with a magnet having p pairs of poles and a shaft to be coupled to a turbomachine body, an inverter converter comprising a control input and outputs each connected to one of the X phases to control the electric motor, a means for measuring the current delivered on the X phases and a position sensor comprising a rotating mobile part mounted integrally on the rotor shaft and a fixed part integral with the stator, the control loop being configured to carry out the method for controlling a turbomachine comprising an electric machine, according to the preceding aspect of the invention with or without the different characteristics of one of the two embodiments described previously, to calculate a command of the inverter converter, the control loop comprising: • a current regulator comprising: • a direct and quadratic current setpoint input, • a quadratic and direct current return input, • a voltage setpoint output, • a calculation unit to carry out a Park and inverse Park transformation comprising: • a direct and quadratic voltage setpoint input, connected to the voltage setpoint output of the current regulator, • a measured current input, connected to the measuring means, • an adjusted angular position input, • a current feedback output for outputting a quadratic return current and a direct return current, the current feedback output being connected to the feedback input of the current regulator, • a control output intended to be connected to the control input of the inverter converter, • an angle error compensator, comprising: • a measurement input intended to be connected to the angle measurement sensor, • a current return input, connected to the current return output of the calculation unit to receive the quadratic return current and the direct return current, • a voltage reference input connected to the voltage reference output of the regulator to receive the direct and quadratic reference voltage, • an adjusted angle output connected to the adjusted angular position input of the calculation unit, • an adjusted angle calculator configured for: • calculate an angular difference value based on the current measurements and the voltage setpoint and the rotor speed, to identify on the one hand the angular error on the direct axis and on the other the angular error on the quadratic axis, • calculate an adjusted angle from the angular difference value based on the comparisons.

[0033] Thus the current regulator being configured to transform a setpoint comprising a quadratic setpoint current and a direct setpoint current received at its setpoint input, into a voltage setpoint comprising a quadratic and direct voltage with an angle at its setpoint output according to the return of quadratic return current and direct return quadratic current received at its return input.

[0034] The computing unit can perform the Park and inverse Park transformation, being configured to: • transform the measured currents received at the measured current input into a quadratic current and a direct current in a Park reference frame, according to the value of the adjusted angle and output this quadratic current and this direct current at its return output, • transform an inverse Park transformation on a quadratic and direct voltage received at the voltage reference input to transmit to the control output a pulse width modulation command to the inverter converter in order to supply the stator windings of the main stator.

[0035] The angle error compensator may be configured to compare, during the period t, the forward voltage of the voltage setpoint to a value of inverse sign, the value being equal to the calculated electrical speed multiplied by a quadratic inductance multiplied by the quadratic return current, and compare the quadratic voltage of the voltage setpoint to a first value equal to the calculated electrical speed of the rotor multiplied by a predetermined value corresponding to the flux of the rotor magnets.

[0036] Another aspect of the invention relates to an electrical machine comprising the rotor, the stator, the position sensor and the inverter converter comprising a control input and outputs each connected to one of the X phases to control the electric motor and the regulation loop.

[0037] The invention and its various applications will be better understood upon reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES

[0038] The figures are presented for information purposes only and in no way limit the invention.

[0039] [Fig-1] shows a representation of a block diagram with functional blocks functionalities of an electric machine allowing the invention to be implemented.

[0040] [Fig.2] shows a schematic representation of a calculation law for the calculation of a adjusted angle to perform a method of automatically calibrating an angular position sensor according to an example of an embodiment of the invention.

[0041] [Fig.3] shows a schematic representation of a method for automatically calibrating an angular position sensor according to an example of a first embodiment of the invention.

[0042] [Fig.4] shows a schematic representation of an automatic calibration process of an angular position sensor according to an example of a second embodiment of the invention. DETAILED DESCRIPTION

[0043] The figures are presented for information purposes only and in no way limit the invention.

[0044] [Fig.l] represents an electrical machine M of a turbomachine, comprising an electric motor 1 comprising a stator comprising X phases, a rotor surrounded by the stator. The electrical machine M further comprises an angular position sensor 2 making it possible to measure the position of the rotor relative to the stator. The electric motor 1 can operate in motor mode and / or in a generator mode. The electrical machine M is a multi-phase synchronous electrical machine of an aircraft thruster.

[0045] The rotor comprises a shaft and an active part mounted on the shaft. The active part of the rotor comprises magnets forming P pairs of poles. The angular position sensor 2 comprises a movable part mounted on the rotor shaft and integral in rotation with the rotor shaft, a fixed part mounted on a fixed part of the electric motor integral with the stator, for example a bearing of the electric motor 1.

[0046] The position sensor 2 can be a resolver or a Hall effect sensor.

[0047] The number X in this example is three, but could be higher for example five or six. The stator therefore comprises in this example three windings forming three phase outputs U, V, W of three phases which are represented in [Fig.l] as coupled in stars but could for example be coupled in delta.

[0048] The electrical machine M further comprises an inverter converter C for supplying the X phases and a regulation loop R making it possible to implement a method for automatically calibrating the angular position sensor 2 relative to the rotor according to a first and a second embodiment explained below.

[0049] The inverter converter C is in this example an inverter with a direct voltage input source, i.e. a DC / AC converter, but could be a converter inverter with alternating voltage input source (i.e. an AC / AC inverter converter) or a reversible inverter converter, i.e. AC / DC and DC / AC in the case where the electrical machine M also allows generator mode. The inverter converter C comprises N outputs, in this case three outputs each connected to one of the corresponding phase outputs U, V, W. The inverter converter C also comprises power inputs, in this case two DC voltage inputs connected to the terminals of a DC voltage bus B. The inverter converter C comprises electronic power switches and a setpoint input, in this case pulse width modulation to power the electronic power switches and thus vary in motor mode the electrical power supply to the stator phases to vary the speed and torque of the rotor of the electric motor 1.

[0050] The electrical machine M further comprises a measuring means 3 for measuring phase currents Iu, Iv, Iw flowing on the phase outputs U, V, W.

[0051] The regulation loop R comprises a regulator 4 comprising a direct current setpoint input Id# and quadratic current Iq#, hereinafter referred to as setpoint I#. The setpoint I# may come from a control unit of the thruster or from several thrusters transmitting the setpoint I# in direct currents Id# and quadratic current Iq# with an angle α to the regulator 4. The regulator 4 comprises a voltage setpoint output V' comprising a direct setpoint voltage Vd' and a quadratic setpoint voltage Vq', the calculations of which are explained below.

[0052] The regulation loop R further comprises a calculation unit 5 using a known mathematical method, namely the so-called direct “Park” transform to go from a three-phase reference frame U; V; W linked to the stator to a two-phase rotating reference frame d; q while also knowing the angular position 0 of the rotor of the electric motor 1 relative to its stator, as well as the inverse Park transform to go from the Park reference frame d; q to the three-phase reference frame U; V; W while also using an angular position of the rotor. The calculation unit 5 comprises a voltage setpoint input connected to the voltage setpoint output of the regulator 4 to receive the voltage setpoint V', an inverse transformation and control calculator, and a control output connected to the control input of the inverter converter C, to transmit the PWM command calculated by the direct transformation calculator, in this case in pulse width modulation.The inverse Park transformation and control calculator is therefore configured to transform the vectors of the direct and quadratic reference voltages Vd', Vq' in a Park plane from an adjusted angle Ocontrol explained in the following, into a PWM pulse width modulation control. The PWM pulse width modulation control includes electrical voltage signals, to control each phase in pulse width modulation and thus generate a balanced three-phase alternating voltage system.

[0053] The calculation unit 5 further comprises measured current inputs receiving the measured phase currents iu, iv, iw, the measured current inputs are connected to the measuring means 3, and a quadratic current return output Iq and direct current Id connected to the return input of the current regulator 4.

[0054] The calculation unit 5 further comprises an adjusted angular position input receiving an adjusted angle Ocontrol and a Park transformer calculator configured to transform the measured currents lu, Iv, Iw according to the adjusted angle Ocontrol received at the adjusted angle input into a component on the quadratic axis of the current vector called in the sequence quadratic current Iq, and the component on the direct axis of the direct current vector Id in a Park plane, also called in the sequence return direct current Id.

[0055] The direct return current Id is therefore in the following the direct component of the stator current in the Park plane calculated from the measured stator currents.

[0056] The quadratic return current Iq is therefore in the following the quadratic component of the stator current in the Park plane calculated from the measured stator currents.

[0057] The current regulator 4 comprises adders 41 (represented by a single adder in [Fig.l]), receiving as non-inverting input the setpoint quadratic current vector Iq# and the setpoint direct current vector Id# and as inverting input the return quadratic current vector Iq as well as the return direct current vector Id.

[0058] At the output of the adders 41, the regulator 4 comprises a line on which circulates the difference between the quadratic current setpoint Iq# and the quadratic return current Iq as well as on another line the direct setpoint current Id# and the direct return current Iq. The two lines are represented by a single line. The current regulator 4 comprises PI (proportional integral) current regulators 42 connected to the difference lines and delivering voltages, respectively the voltages Vq' and Vd'.

[0059] The control loop R further comprises an angle error compensator 6 comprising a measurement input connected to the angle measurement sensor 2 and a signal processing unit 61 making it possible to transform the electrical signals received from the position sensor 2 into an angle measurement, called the measured angle Osensor.

[0060] The angle error compensator 6 further comprises a current return input, connected to the return current output of the calculation unit 5 receiving the quadratic return current Iq and the direct return current Id, a voltage setpoint input connected to the voltage setpoint output of the current regulator 42 receiving the setpoint voltage V' and an adjusted angle output Ocontrol connected to the adjusted angle input of the calculation unit 5. The angle error compensator 6 comprises an adjusted angle calculator 62 (a calculation unit) for determining the adjusted angle Ocontrol from the direct voltage Vd', the quadratic voltage Vq', the quadratic return current Iq, the direct return current Id, an electrical speed of the rotor œe.iec calculated from the calculated rotation speed N by the measurements of the measured angles Orotor sensor and a predetermined number corresponding to the number P of pole pairs on the rotor, a quadratic inductance Lq.

[0061] [Fig. 2] represents a block diagram of the calculation of the angle error compensator 6 and [Fig. 3] represents a schematic representation of control steps of a method for automatic calibration of the angular position sensor 2 according to an example of a first embodiment of the invention. The automatic current loop steps are not shown in this [Fig. 3] but are described below.

[0062] In this first embodiment, the rotor of the electric motor 1 of the electric machine M can be decoupled from the body of the turbomachine to operate at no load.

[0063] The method for automatically setting the angular position sensor 2, according to this embodiment, is carried out at each start of the electric machine M, starting by decoupling the rotor from the body of the turbomachine. The start may be initiated, for example, by a setpoint I# received from a control unit of the turbomachine. The method comprises, in this example, a first part for identifying an error in the position sensor by the speed El, comprising a first control step 1E consisting of forcing the setpoint I# received to zero, i.e. the quadratic current Iq# and direct current Id# to 0, and a second step 2E consisting of calculating a command for the inverter converter C in order to rotate the induced rotor of the electric motor 1 at a predetermined speed Nn in a constant manner.The first part of error identification of the position sensor by the speed El is repeated several times by changing the predetermined speed Nn, here in this embodiment by incrementing it, the index n represents the number of times of the part of error identification of the position sensor by the speed El during this process.For example the first predetermined speed Nn=i is equal to 100 rpm and at the second time of the error identification part of the position sensor by the speed El, the PWM command is calculated so that the rotation speed N of the rotor goes from the first predetermined speed Nn=i to the second predetermined speed NThe method therefore comprises a first current loop step consisting of calculating a voltage setpoint V' and transforming it into a PWM pulse width modulation command by the calculation unit 5 which transmits it to the inverter converter C which supplies by transforming the direct current of the DC bus into three alternating currents (modulated) read Iv, Iw of a three-phase system, for the three phases UVW in this example of the stator of the electric motor 1 so that the rotor at a rotation speed N = at the predetermined rotation speed Nn. .

[0064] The part of identifying the error of the position sensor by the speed El of the method further comprises a second current loop step, of measuring the currents of the phases lu, Iv, Iw by the current measuring means 3 and of transforming in a Park reference frame, into quadratic return current Iq and into direct return current Id, taking into account the value of an adjusted angle ôsensor calculated, explained subsequently.

[0065] The error identification part of the position sensor El by the speed comprises a third step 3E comprising a sub-step not shown consisting of calculating an electrical speed of the rotor (üelecen) by multiplying the number p of pairs of poles of the rotor (predetermined) by the predetermined mechanical speed.

[0066] The third step 3E comprises a sub-step not shown of calculating the voltage setpoint V' by the regulator during a period t, as a function of the setpoint current comprising a forcing of the direct setpoint current equal to 0 and from the quadratic return current Iq and the direct return current id, the return currents measured and transformed by the Park transformation at this constant predetermined rotation speed N„.

[0067] The third step 3E further consists of comparing during the period t the direct voltage Vd' of the voltage setpoint V' with a value of inverse sign, the value being equal to the calculated electrical speed Meie.c multiplied by a quadratic inductance Lq multiplied by the quadratic current Iq, and comparing the quadratic voltage Vq' of the voltage setpoint V' with a first value equal to the calculated electrical speed of the rotor Meiec multiplied by a predetermined value corresponding to the flux of the rotor magnets.

[0068] The adjusted angle calculator 62 can also make it possible to calculate a first value equal to the product of the electrical speed calculated with the quadratic inductance Lq and the quadratic current Iq.

[0069] The adjusted angle calculator 62 further calculates a second value of the calculated electrical speed welec multiplied with a predetermined value Vf corresponding to the flux of the rotor magnets. The adjusted angle calculator 62 shown in FIG. 2 allows the calculation of the angular difference value A 6 between the adjusted angular position Ocontrol and the measured angular position Osensor by the position sensor. In particular, the adjusted angle calculator 62 identifies the offset direct axis angular angle (measured direct angle error) relative to the component on the direct axis of the return current vector as well as the offset quadratic axis angular angle S(J (measured direct angle error) relative to the component on the quadratic axis of the return current vector.

[0070] In this example, the adjusted angle calculator 62 shown in Figure 2 calculates so that the offset direct axis angular angle SJ and the offset quadratic axis angular angle tf / are each equal to zero.

[0071] The adjusted angle calculator 62 presented in [Fig.2] therefore comprises in this example a summer 620 to carry out the comparison by receiving at a non-inverting input a value 0 and another non-inverting input the sum of the direct voltage Vd' and the value of the first product. As in this embodiment, the instruction is to have the direct return current Id and the quadratic return current Iq equal to zero, the adjusted angle calculator 62 can also compare Vd' to the value 0 since the product described above is equal to zero. The quadratic inductance Lq does not vary unless the quadratic return current Iq is very large, as in this example, the quadratic return current Iq = 0 or close to zero if error, the quadratic value can here be a predetermined value.

[0072] The adjusted angle calculator 62 therefore further comprises in this example a summer 621 for carrying out a comparison by receiving as a non-inverting input a value 0 and as a non-inverting input the sum of the quadratic voltage Vq' and the second calculated value. According to another example not shown, the adjusted angle calculator 62 comprises a summer comprising as a non-inverting input the quadratic voltage Vq' and as an inverting input the second calculated value.

[0073] The adjusted angle calculator 62 can therefore determine a rotor position angle called the estimated angular position Ocontrol and therefore an angular difference value A 0 between the adjusted angular position Ocontrol and the measured angular position Osensor by the position sensor.

[0074] Indeed, we know that the couple ? em = 2 P [ Vf iq " (Lq - îqîq j ' as the torque is at zero (no-load machine therefore Iq = 0), in the event of angular error measured by position sensor 2, and assuming the voltage drop across the stator resistors to be negligible, the direct voltage Vd' and quadratic voltage Vq' are equal to the following formula: Vd* -^eleC(LVq + VpM^9) Vq * Welec ((Vf + Ldid) - Lqiq A 0) • Since Id has been forced to 0, we can determine A 0 A — 4 [ [kp + ~sj ( kp + T j Sq

[0075] with Ed = (ûetecLqiq + Vd and = - a)eiec(Pf+ Vq and kp, k( are the gains of the proportional integral type corrector used to cancel the error ( X#-X ) and ( X# -X 4 \ dd 7 \ 9 4 7

[0076] Ld is the direct component of the stator inductance in the Park plane,

[0077] Lq is the quadratic component of the stator inductance in the Park plane. Regardless of the regulator used, the direct current component setpoint will come from the optimal direct current component mapping.

[0078] Means for determining the quadratic component Lq of the inductance of the synchronous electric motor 1 can use values ​​of the components of the direct return current Id and of the quadratic return current lq and maps of the electric motor 1.

[0079] The part of identifying the error of the position sensor by the speed El of the method of automatically calibrating the angular position sensor 2 therefore comprises a fourth step 4E consisting of calculating the angular difference value AB between the adjusted angle Bcontrol and the measured angle ^sensor by the position sensor 2.

[0080] It can thus be seen in [Fig.2] that in this example, the adjusted angle calculator 62 further comprises a first and a second proportional-integral type corrective regulator, 624, 625 at the output of the adders 620, 621 respectively of the first and second comparison, as well as a adder 626 and a multiplier 627 by 0.5 to average the difference values.

[0081] The error identification part of the position sensor by the speed El therefore comprises a fifth step 5E consisting of calculating the adjusted angle Bcontrol by adding the angular difference value A 0 to the measured angle ^sensor.

[0082] The angle error compensator 6 therefore comprises a summator 63 of the measured angle ^sensor and the angular difference value AB to obtain the adjusted angle 7control.

[0083] Thus, at the output of the angle error compensator 6, the calculation unit 5 receives the adjusted angle ^control to carry out the direct and inverse Park transformations described previously in the current loop.

[0084] In this example of this embodiment, the angle error compensator 6 comprises a memory 65 recording the angular difference value A 8 as well as the corresponding predetermined speed Nn. The method therefore comprises a step of recording the angular difference value AB and the predetermined value of the corresponding rotation speed Nn, here in this example Nn= K * n or 100 rpm * n.

[0085] Furthermore, in this example of this embodiment, the method for automatically setting the angular position sensor 2 verifies that the predetermined speed Nn is not equal to half the maximum speed NMax of the rotor of the electric motor 1 and in that the method repeats the steps described previously of the error identification part of the position sensor by the speed El by incrementing the rotational speed N in a predetermined manner. Here in this example, the error identification part of the position sensor by the speed El is repeated several times by reducing each time the predetermined rotational speed Nn, here in this case by the value K, for example 100 rpm. The number of times (the index n) where the error identification part of the position sensor by the speed El is carried out is therefore equal to the max speed NMax divided by K.

[0086] The value of the rotation speed N is therefore increased by 100 rpm, until the latter reaches a predetermined value NMax. In this example, the predetermined value NMax is equal to half of the predetermined maximum rotor speed. Carrying out the error identification part of the position sensor by the speed El at different speeds by increasing it makes it possible to refine the calculated angular difference.

[0087] The compensator 6 comprises a summer 63 receiving as a non-inverting input the angle measurement Osensor delivered by the processing unit 61 and, as a non-inverting input, the angular difference A0 making it possible to calculate an estimated calculated angle Ocontrol and transmit it to the calculation unit 5.

[0088] The current loop steps described above are simultaneous because they are in a closed control loop.

[0089] In this example of this embodiment, the method for automatically calibrating the angular position sensor 2 further comprises a part for identifying an error of the position sensor by the torque E2 when the predetermined speed is equal to or greater than the predetermined value NMax.

[0090] The error identification part of the position sensor by the torque E2 requires coupling a load on the rotor shaft of the electric machine 1.

[0091] The error identification part of the position sensor by the torque E2 comprises a first step 6E of calculating a PWM command of the inverter C to supply the x phases in order to rotate the induced rotor at a speed for the calculation by the predetermined torque Ni in a constant manner, the command being calculated by the calculation unit by calculating the speed according to the position measurements by the position sensor 2.

[0092] This part of error identification of the position sensor by the torque E2 includes the steps of the current loop described previously: the measurement of the currents lu, Iv, Iw of the phases and transforming them in a Park reference frame, into quadratic current Iq and into direct current Id, taking into account the calculated adjusted angle, but also the calculation of an electrical speed of the rotor ^L / « by multiplying the number of pairs of poles by the predetermined speed.

[0093] This part of identifying the error of the position sensor by the torque E2 comprises a step 7E of imposing at the predetermined speed Ni at the input of the re current regulator 4, a quadratic current with setpoint Iq# at a first predetermined value Ik different from zero, and a direct current with setpoint Id# at a zero value 0, from the measured currents transformed by the Park transformation.

[0094] The error identification part of the position sensor by the torque E2 comprises a step 8E consisting of comparing at the second predetermined speed Ni in a constant manner, the value of the direct voltage of reference Vd' with the first value equal to ■ The quadratic inductance Lq can be calculated or determined or be pre determined. The angle error compensator 6 can use the adjusted angle calculator 62 to carry out this step in a similar way to the part of calibrating the position sensor by the speed EL Thus, step 8E further comprises the comparison with the second speed predetermined in a constant way, that at the output of the current regulator, the value of the quadratic voltage of reference Vq' is equal to the second calculated value tüejec* <p

[0095] The error identification part of the position sensor by the torque E2 comprises a step 9E of calculating the angular difference A0 with the adjusted angle calculator 62 while the rotor rotates at the second predetermined speed Ni constantly according to these comparisons, and recording the angular difference A0 with the predetermined speed Ni.

[0096] The automatic calibration method repeats a predetermined number of times the part of checking the calibration of the angular position sensor 2 by the current by incrementing in predetermined steps, the predetermined value Ik of the set quadratic current Iq# until the return quadratic current Iq is equal to a final predetermined value corresponding in this example to the predetermined maximum return quadratic current value Iqmax. This makes it possible to refine the calculated angular difference A0 and record it at each part of checking the calibration of the angular position sensor 2 by the current.

[0097] As the angular difference A0 allows to have an adjusted current Ocontrol in the current loop, the angular difference A0 tends towards 0 and is verified, if the angular difference A0 is different from 0 when the quadratic return current Iq is equal to a final predetermined value, the automatic calibration method repeats a predetermined number of times the part of checking the calibration of the angular position sensor 2 by the current from the beginning, otherwise it goes into a normal mode of operation of the PWM control of the electric motor 1.

[0098] The method for controlling an electrical machine of a turbomachine in so-called normal mode comprises the steps of the current loop, using the direct current setpoint Id# and quadratic Iq# and the quadratic return current Iq and the direct return current Id calculated by the calculation unit 5 by the mathematical method of the Park transform of the measured currents read, Iv, Iw, according to an angle B including the measurement of an angle Osensor added to a value according to a recorded angular difference A 9.

[0099] [Fig.5] represents an example of calculation of the angle B which is equal to the value of the measured angle Osensor added to an average angular difference AOmoy calculated from the angular difference values ​​AO recorded in the memory 65.

[0100] [Fig.4] represents a second embodiment of the method for automatically calibrating the angular position sensor 2. Unlike the first embodiment, this method makes it possible to automatically calibrate the angular position sensor 2 of the rotor of the electric motor 1, the rotor of which is coupled to a body of a turbomachine. The body causes an inertial torque and therefore the electric machine does not actually rotate at no load.

[0101] In this second embodiment of the method for automatically calibrating the angular position sensor 2, the method comprises, as in the first embodiment, a part for identifying an error in the position sensor by the speed 1E' except that it is different in that the predetermined rotation speed N'n is the max rotation speed Nmax and that each time the method repeats the part for identifying an error in the position sensor by the speed 1E', the value of the predetermined speed N'n is reduced, for example here, by a constant k.

[0102] In this example of this embodiment, the first step 1E' is the calculation of a PWM command of the inverter converter C) in order to rotate the induced rotor at a predetermined speed N'n constantly, i.e. the first time at the first predetermined speed N'i (the index n being = 1).

[0103] The second step 2E' is the step of calculating a voltage setpoint V' during a period t to obtain a direct current value Id equal to 0 from the quadratic return current Iq and the direct return current Id transformed by the Park transformation, is carried out by imposing a quadratic current setpoint Iq# equal to zero and a direct current setpoint Id# equal to zero.

[0104] In other words, in this second embodiment, the first step 1E' and second step 2E' correspond respectively to the second step and the first step of the first embodiment, except in that the first predetermined speed N'i is the maximum speed Nmax in this embodiment whereas conversely in the first embodiment, the first predetermined speed Ni is the lowest of the predetermined speeds Nn.

[0105] The third step 3E', the fourth step 4E', and the fifth step 5E' of this second embodiment correspond respectively to the third comparison step 3E, the fourth step 4E of calculating the angular difference value A 0, and the fifth step 5E of calculating the adjusted angle dcontrol of the first embodiment. lization.

[0106] As explained previously, in this embodiment, the error identification part of the position sensor by the speed El' is repeated several times by reducing each time the predetermined rotation speed N'n, here in this case by a value k, for example 200 rpm. The number of times the error identification part of the position sensor by the speed El' is carried out is therefore equal to the max speed divided by k.

[0107] When the predetermined rotation speed N'n equals zero (the index n therefore has an equal value Nmax / k), a normal mode control method of the electric motor 1 can begin for calculating the control of the inverter converter C to supply the phases of the stator of the electric motor 1 according to the received direct current Id# and quadratic Iq# setpoint and the values ​​of quadratic return current Iq and direct return current Id calculated by the calculation unit 5 by the mathematical method of the Park transform of the measured currents lu, Iv, Iw, according to an angle B comprising the value of the measured angle Osensor added to a correction value according to a recorded angular difference A 0.The correction value may be, as in the first embodiment, the average of the angular difference values ​​A0moy recorded in the memory 65 but may be, according to another example, the calculated angular difference value A 0 recorded corresponding to the predetermined speed Nn recorded closest to the measured instantaneous speed N. .

[0108] Unless otherwise specified, the same element appearing in different figures has a single reference.

Claims

Claims

1. Method for automatically calibrating an angular position sensor (2) measuring an angle corresponding to the mechanical angular position of an induced rotor with a magnet forming P pairs of poles relative to a stator comprising X phases of a multi-phase synchronous electric motor (1) of an electric machine (M) of an aircraft thruster, the position sensor (2) comprising a movable part mounted on the induced rotor and a fixed part relative to the stator, the method using a regulation loop (R) to transmit a command (PWM) to an inverter converter (C) of the electric machine (M) to supply the phases of the stator,the control loop (R) comprising a current regulator (4) comprising a setpoint input for receiving a current setpoint (I#) comprising a quadratic setpoint current (Iq#) corresponding to a component on the quadratic axis of the setpoint current vector and a direct setpoint current (Id#) corresponding to a component on the direct axis of the setpoint current vector, for delivering a voltage setpoint (V') comprising a quadratic voltage (Vq') and direct voltage (Vd') with an angle (a), a calculation unit (5) for calculating a direct Park and inverted Park transformation, making it possible to transform the voltage setpoint (V') into control (PWM) of the inverter converter (C), an angle error compensator (6) for identifying an angle error and compensating for it, the method being characterized in that it comprises a part for identifying the error of the position sensor by the speed (El, El'), comprising the following steps:, - calculate (2E, 1E') the command (PWM) of the inverter converter (C) in order to rotate the induced rotor at a predetermined mechanical speed (Nn) constantly, - measure the phase currents (lu, Iv, Iw) and transform them in a Park frame, into quadratic return current (Iq) and direct return current (Id), taking into account a calculated adjusted angle (Bcontrol), - calculate an electrical speed (qj) of the rotor by multiplying the number p of pole pairs by the predetermined mechanical speed (Nn), calculating (2E, 1E') a voltage setpoint (V') during a period t in which the rotor rotates at the predetermined mechanical speed (jy) by imposing a forward current value equal to 0 on the current regulator, Identifying a shift error of the direct component on its direct axis by comparing (3E, 3E') during the period t the forward voltage (Vd') of the voltage setpoint (V') to a value of inverse sign, the value being equal to the calculated electrical speed (MeiPC) multiplied by a quadratic inductance (Lq) multiplied by the quadratic return current (Iq), and Identify a quadratic component offset error (¾) on its quadratic axis by comparing, during the period t, the quadratic voltage (Vq') of the voltage setpoint (V') with a first value equal to the calculated rotor electrical speed ^eiec) multiplied by a predetermined value (^ / ) corresponding to the flux of the rotor magnets, calculate (4E, 4E') an angular difference value (A 0) from these identified direct and quadratic angular offset errors, determine (5E, 5E') at the period t, an adjusted angle (Ocontrol) according to the calculated angular difference value (A 0), and the value of the measured angle (^sensor), transmit the quadratic return current (Iq) and the direct return current (Id) according to the adjusted angle (Ocontroï) to the current regulator (4).

2. Method for automatically calibrating an angular position sensor (2) according to the preceding claim in which the calculation of the angular difference value (A 8) is calculated according to this formula: A3- 4 [ + with 8j — cû^Lqiq + Vj corresponding to the comparison to identify the angular error of the direct component and - weiec <Pf + Vq correspondant à la comparaison pour identifier l’erreur angulaire de la composante quadratique, Lq est l’inductance quadratique, kp, k{ sont les gains du correcteur proportional-integral type used to cancel the error between respectively the setpoint quadratic current (Iq#) and the return quadratic current (Iq) and the setpoint direct current (Id#) and the return direct current (Id).

3. Method for automatically calibrating an angular position sensor (2) according to one of the preceding claims, in which the rotor is coupled to a shaft of the propeller comprising an inertial torque, in which during the period t, the value of the direct voltage (Vd') is imposed to be equal to 0.

4. Method for automatically calibrating an angular position sensor (2) according to one of the preceding claims, in which the rotor is coupled to a body of a turbomachine, and in that the part of calibrating the position sensor (2) by the speed (El') is repeated a number X times at different predetermined speeds and in that each time that a part of calibrating the position sensor by the speed is carried out, the angular difference value (A ff) between the adjusted angle (^control) and the measured angle (^sensor) and the corresponding predetermined mechanical speed (N') is recorded.

5. Method for automatically calibrating an angular position sensor (2) according to claim 4, in which the first predetermined mechanical speed (N'i) is the highest speed of the predetermined speeds (Nmax), for example the maximum rotational speed of the rotor, and in that the method restarts the part of calibrating the position sensor by the speed (El') by decreasing the mechanical speed of the rotor (N') according to a predetermined difference (k) until the rotor stops rotating.

6. Method for automatically calibrating an angular position sensor (2) according to one of claims 1 to 2, in which the electrical machine can rotate at no load, for example by disengaging from a turbomachine shaft, the method comprises: - in the step of calculating a command of the inverter converter (2E) of the error identification part of the position sensor by the speed (El), the predetermined mechanical rotation speed (N) is incremented in a predetermined manner each time the method restarts the part

7. position sensor error identification by speed (El), - in the error identification part of the position sensor by speed (El), a step of recording each angular difference value (A 0) between the calculated adjusted angle value (ücontrol) and the measured angle (^sensor). Method for automatically calibrating an angular position sensor of a multi-phase electrical machine according to claim 6, further comprising after the part of calibrating the position sensor by the speed (El) a part of verifying the calibrating of the angular position sensor (2) by the current (E2) comprising the following steps: • calculate (6E) a command of the inverter converter (C) to supply the X phases in order to rotate the induced rotor at a second predetermined mechanical speed (Ni) constantly, • measure the currents (iu, Iv, Iw) of the phases and transform them in a Park reference frame, into quadratic return current (Iq) and direct return current (Id), taking into account the value of the adjusted angle (Qcontrol) calculated, • calculate an electrical speed of the rotor (co_elec) by multiplying the number (P) of pairs of poles by the second predetermined mechanical speed (Ni), • impose (7E) at the predetermined mechanical speed (Ni), at the input of a current regulator (4), a quadratic setpoint current (Iq#) at a first predetermined value (Ik) different from zero, and a direct setpoint current (Id#) at a zero value, • determine the quadratic inductance (Lq), • compare (8E) to the second predetermined mechanical speed constantly, that at the output of the current regulator (4), the value of the direct set voltage (Vd') is equal to a value of inverse sign, the value being equal to the product of the calculated electrical speed (co_elec), the quadratic inductance (Lq) and the quadratic return current (Iq), • compare (8E) to the second predetermined mechanical speed constantly, as at the output of the current regulator (4),

8.

9. the value of the setpoint quadratic voltage (Vq') is equal to the product of the calculated electrical speed (co_elec) times the predetermined value corresponding to the flux of the magnets of the rotor, • calculate (9E) the angular difference (A 0) between the calculated adjusted angle (Ocontrol) and the measured angle (^sensor), while the rotor rotates at the second predetermined mechanical speed (N;) constantly according to these comparisons, and record it, • in which the automatic calibration method repeats a predetermined number N of times the part of checking the calibration of the angular position sensor (2) by the current (E2) by incrementing each time the predetermined value (Ik) of the setpoint quadratic current (Iq#) in a predetermined manner until a predetermined maximum return quadratic current value (Iqmax) is obtained. Method for controlling an electric machine of a turbomachine, in which at each start of the electric machine, the method comprises the method of automatically setting an angular position sensor (2) according to any one of the preceding claims, and in which the method further comprises a normal mode of control of the electric motor (1) comprising a step of - calculation of the control of the inverter converter (C) to supply the phases of the stator of the electric motor (1) according to the setpoint (I#) received from the values ​​of quadratic return current (Iq) and direct return current (Id), the calculation unit (5) carrying out the calculations by the mathematical method of the direct or inverse Park transform, according to an angle B comprising the measured angle (Osensor) added to a correction value according to a recorded angular difference (A 0). Method for controlling a turbomachine comprising an electric machine, according to the preceding claim in which the correction value can be equal to the angular difference value (A 0) calculated recorded corresponding to the predetermined mechanical speed (Nn) recorded closest to the measured instantaneous mechanical speed (N).

10. Method for controlling a turbomachine comprising an electric machine, according to claim 8, further comprising a step of calculating an average (A Gmoy) of the recorded angular difference values ​​(A 0) and in which the correction value is equal to the average (A Omoy) of the recorded angular difference values ​​(A 0).

11. Control loop (R) of an electric machine (M) of a turbomachine for calculating an inverter converter command (C) for an electric motor (1) comprising an inductor stator having stator windings forming X phases, an induced magnet rotor having p pairs of poles and a shaft for being coupled to a turbomachine body, an inverter converter (C) comprising a control input and outputs each connected to one of the X phases (u, v, w) for controlling the electric motor (1), a measuring means (3) for measuring the current delivered on the X phases (u, v, w) and a position sensor (2) comprising a rotating movable part mounted integrally on the rotor shaft and a fixed part integral with the stator, the control loop (R) being configured to carry out the method for controlling a turbomachine comprising an electric machine, according to one of claims 8 to 10, to calculate a command (PWM) of the inverter converter (C),the control loop (R) comprising:, 1. a current regulator (4) comprising: a. a direct and quadratic current setpoint input, b. a quadratic and direct current feedback input, c. a voltage setpoint output, 2. a calculation unit (5) for carrying out a Park and inverse Park transformation comprising: a. a direct and quadratic voltage setpoint input, connected to the voltage setpoint output of the current regulator (4), b. a measured current input, connected to the measuring means (3), c. an adjusted angular position input, d. a current feedback output for outputting a quadratic return current (Iq) and a direct return current (Id), the current feedback output being connected to the feedback input of the current regulator (4), e. a control output intended to be connected to the control input of the inverter converter (C), 3. an angle error compensator (6), comprising: a. a measuring input intended to be connected to the angle measuring sensor (3), b. a current return input, connected to the current return output of the calculation unit (5) to receive the quadratic return current (Iq) and the direct return current (id), c. a voltage setpoint input connected to the voltage setpoint output of the regulator (42) to receive the direct and quadratic setpoint voltage (Vd', Vq'), d. an adjusted angle output connected to the adjusted angular position input of the computing unit (5) e. an adjusted angle calculator (62) configured to:

4. calculate an angular difference value (A 0) based on the current measurements and the voltage setpoint and the rotor speed, to identify on the one hand the angular error on the direct axis and on the other the angular error on the quadratic axis, 5. calculate an adjusted angle (dcontrol) from the angular difference value (A ff).

12. Electrical machine (M) comprising the rotor, the stator, the position sensor (2) and the inverter converter (C) comprising a control input and outputs each connected to one of the X phases to control the electric motor (1) and the control loop (R) according to the preceding claim.

Images