Method and device for estimating a speed of a rotor of a brushless motor by speed-dependent averaging

Abstract

The invention relates to a method for estimating a speed (ωi) of the rotor of a brushless motor (10), the method comprising updating a previous estimate (ωi-1) of the speed, the update being triggered by receipt of signals from a set of position sensors (11) positioned on the motor (10), and the method comprising the steps of: - detecting (S1) an angular position (θi) of the rotor from the signals, and determining a time (ti) of receipt of these signals; - determining and storing (S2) a new position difference (Δθi) and a new time interval (Δti); - determining (S3) a speed range as a function of a previous speed range and of the previous estimate of the speed; - estimating (S4) a speed ((ωi) as an average value over a succession of stored position differences and time intervals, the number of which is an increasing function of the speed range.

Description

DESCRIPTION TITLE: Method and device for estimating the speed of a brushless motor rotor by speed-dependent averaging TECHNICAL FIELD

[0001] The domain is that of brushless motor control in an aeronautical context, particularly when mounted on an aircraft. The proposal specifically concerns the estimation of the motor's rotor speed in order to adapt the motor's control input. TECHNOLOGICAL BACKGROUND

[0002] In the aeronautical field, brushless motors play a crucial role due to their high reliability, energy efficiency, and ability to operate without complex mechanical maintenance. These characteristics make them a preferred choice for critical systems such as aircraft electric propulsion or auxiliary systems (e.g., electric actuators).

[0003] A "brushless" motor (or BLDC for "Brush Less Direct Current" in English, i.e., direct current and without brushes) operates from 3 variable voltage sources which generate, by means of windings located on the stator, a rotating magnetic field capable of driving a rotor.

[0004] The rotational speed of the rotor is a function of the sequence of voltages applied to these windings.

[0005] A control system is therefore planned to generate instructions to a modulator in charge of generating the voltage signals applied to the different windings in order to drive the rotor to the desired rotational speed.

[0006] These instructions depend on a target speed, usually determined from an actuator on the control panel of an aircraft pilot, but also on signals obtained from the engine by a feedback loop.

[0007] Indeed, the actual rotational speed of the rotor may differ from the set speed for various reasons, including uncertainties related to load variations, environmental conditions, etc.

[0008] This feedback loop is therefore necessary to regulate the rotor speed and ensure that it corresponds precisely to the set speed.

[0009] One possibility is to use absolute position sensors mounted on the engine to measure the rotor's position in real time, allowing the speed to be immediately deduced. However, such absolute position sensors present cost, size, and weight constraints that are not well-suited to a system mounted on an aircraft.

[0010] Therefore, less expensive, less bulky and lighter sensors, such as digital Hall effect sensors, are generally preferred.

[0011] These sensors, however, do not provide an absolute rotor position, but rather a discrete position that depends on the number of sensors located around the stator. Therefore, an additional step is required to accurately estimate the rotor speed, separate from the discrete rotor positions provided by Hall-effect sensors.

[0012] Furthermore, it is crucial that this rotor speed estimate respects the real-time constraints of the motor control system. Indeed, excessively long and inaccurate calculations would disrupt the effect of the feedback loop on this control system, such that the commands provided to the modulator would no longer allow the rotor speed to be regulated to the setpoint speed.

[0013] It is therefore important to estimate the rotor's rotational speed with constraints of accuracy and response time. DESCRIPTION OF THE INVENTION

[0014] The invention aims to improve the state of the art. In particular, the problem that the present proposal seeks to solve is to estimate the rotational speed of a rotor of a "brushless" type motor, while simultaneously respecting constraints of accuracy and response time.

[0015] It is proposed to perform this estimation of the speed by an averaging dependent on the speed itself from successions of rotor position measurements provided by Hall effect type sensors.

[0016] More specifically, a method is proposed for estimating the rotor speed of a brushless motor, comprising updating a previous estimate of said speed, said update being triggered by receiving signals from a set of position sensors positioned on said motor and comprising steps, implemented by a speed estimation device, of: detection of a discrete angular position of said rotor from said signals, and determination of a reception time of said signals; determination and storage of a new position deviation and a new time deviation relative to, respectively, a previous discrete angular position of said rotor, and a previous time of reception of said signals; determination of a speed range from a set of speed ranges, each delimited by a lower threshold and an upper threshold, as a function of a previous speed range and said previous estimation of said speed; estimation of a speed as an average value over a succession of stored position deviations and time deviations, the number of which is an increasing function of said speed range.

[0017] According to preferred embodiments, the invention comprises one or more of the following features which can be used separately or in partial combination with each other or in total combination with each other: said speed is estimated to be zero when the determined speed range is the lowest speed range of said assembly; said set of speed ranges includes a lower range corresponding to an upper threshold of preferably 20 rad / s, an intermediate range and an upper range, corresponding to a lower threshold of preferably 50 rad / s; said number of said successions of deviations is 2 for said intermediate range, and 6 for said upper range; said determination of a speed range includes a change from said previous speed range when said previous estimate of said speed is lower or higher than a lower, respectively upper, threshold of said previous speed range, reduced, respectively increased, by a margin, said margin being preferably around 5%; said motor is a brushless DC motor

[0018] Another object relates to a method of controlling a motor comprising a step of determining a setpoint allowing to generate a set of control voltages said motor, said setpoint being determined as a function of a setpoint speed and a speed estimated according to an estimation method as previously described.

[0019] Another aspect concerns a device for estimating the rotor speed of a brushless motor, adapted to perform an update of a previous estimate of said speed, said update being triggered by the reception of signals from a set of position sensors positioned on said motor, said speed estimation device being further adapted to: detect a discrete angular position of said rotor from said signals, and determination of a reception time of said signals; determine and store a new position deviation and a new time deviation relative to, respectively, a previous discrete angular position of said rotor, and a previous instant of reception of said signals; determine a range of speeds from a set of speed ranges, each delimited by a lower threshold and an upper threshold, as a function of a previous speed range and said previous estimate of said speed; estimate a speed as an average value over a succession of stored position and time deviations, the number of which is an increasing function of said speed range.

[0020] Another aspect concerns a motor control system comprising a device for estimating the speed of the rotor of said motor as previously described, and a control chain adapted to determine a setpoint to a modulator adapted to generate a set of control voltages for said motor, said setpoint being determined as a function of a setpoint speed and said speed estimate.

[0021] Another aspect concerns an aerial vehicle, or aircraft, comprising at least one control system as previously described, said set speed being provided by a flight control actuator. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Other aspects, objectives, advantages, and features of the invention will become clearer upon reading the following detailed description of preferred embodiments thereof, given by way of non-limiting example, and made with reference to the accompanying drawings in which: Figure 1 schematically represents an example of a "brushless" type motor; Figure 2 schematically represents a motor control system according to one possible embodiment; Figures 3a, 3b, 3c, 3d illustrate experimental results of estimating the speed of a rotor according to different embodiments; Figure 4 represents an illustrative flowchart of a process according to one embodiment. DETAILED DESCRIPTION OF SPECIFIC METHODS OF IMPLEMENTATION

[0023] The rotor speed estimation method and device are applicable to aerial vehicles or aircraft. These aerial vehicles may be piloted or unpiloted, such as drones or Unmanned Aerial Vehicles (UAVs). They can include airplanes, but also helicopters.

[0024] The rotor speed estimation method and device are particularly applicable to a brushless motor. An example of such a motor is shown in Figure 1.

[0025] A motor 10 consists of a stator 13 and a rotor 12. The stator 13 is equipped with a set of coils, or windings, 14i, 142, 14s, 144, 14s, 14e, regularly distributed around the perimeter of the stator.

[0026] These coils are powered by electrical voltage sources (not shown), which are phase-shifted, generating a rotating magnetic field within the cavity formed by the stator armature. The voltages vary in amplitude and frequency, depending on the required speed and torque.

[0027] Groups of coils, generally pairs of diametrically opposed coils, can be called "phases" by misuse of language, because they are powered by the same voltage signal with a phase shift.

[0028] The rotating magnetic field causes the rotor 12 to rotate with a speed that depends directly on the frequency of the signals applied to the coils.

[0029] In the example in Figure 1, 3 voltage sources are provided, whose signals are offset by 120°, each supplying two diametrically opposed coils (or "phases"), respectively 14i and 144, 142 and 14s, 14s and 14e.

[0030] Furthermore, position sensors Hi, II2, II3, H4, Ils, lie are positioned on motor 10, more precisely in the air gap between the coils.

[0031] These sensors can, for example, be digital Hall effect sensors.

[0032] A digital Hall effect sensor is an electronic device that detects the presence of a magnetic field and provides a binary output (ON / OFF, or 1 / 0) depending on the intensity or orientation of the magnetic field relative to a predetermined threshold.

[0033] When a conductor or semiconductor passes through a magnetic field perpendicular to its current, a voltage (Hall voltage) is generated perpendicular to the magnetic flux and the current. This voltage is a function of the magnetic field strength.

[0034] More precisely, when an electric current flows through a conductor or semiconductor and a magnetic field is applied perpendicular to this current, a force is exerted on the moving charges within the material. The moving electric charges (electrons or holes) experience a Lorentz force due to the magnetic field. This deflection of the charges creates an accumulation of positive charges on one side of the material and negative charges on the other, generating a transverse potential difference (that is, perpendicular to the current). The accumulation continues until the electrostatic force (due to this potential difference) balances the Lorentz force. At this point, a stable transverse voltage, called the Hall voltage, is measured.

[0035] The Hall voltage is the result of the deflection of electric charges by the Lorentz force when moving charges pass through a magnetic field. This voltage is directly related to the magnetic field strength, making it suitable for use as a sensor.

[0036] In a digital sensor, this Hall voltage is compared to a predefined threshold using a detection circuit (such as an integrated comparator). If the magnetic field density exceeds this threshold, the sensor provides a high logic signal (1). Otherwise, it provides a low logic signal (0).

[0037] The voltages applied to the coils are supplied by a modulator controlled by a control system.

[0038] This control system includes a control chain adapted to determine a setpoint to this modulator so that it can generate a set of control voltages for each coil, or more precisely for each group (usually a pair) of coils (corresponding to a "phase" of the motor).

[0039] The angular position of the rotor of a brushless motor can be determined from the signals of Hall effect sensors by exploiting their configuration and the sequence of transitions they generate.

[0040] The Hall sensors detect the proximity of the magnetic poles of the rotor 12. When the rotor is rotating, each Hall sensor Hi, II2, II3, H4, Ils, lie goes through "high" (1) or "low" (0) states, depending on the magnetic polarity encountered.

[0041] In an example with 3 Hall sensors, the possible binary combinations are 2 3=8. However, only 6 distinct combinations are actually used for a three-phase BLDC motor (the remaining two are invalid and should not occur). Each Hall signal combination corresponds to a specific angular position of the rotor in an electrical cycle.

[0042] Thus, in a motor with N pairs of coils, each combination of Hall signals corresponds to an angular interval of: where 6xN is the total number of Hall transitions per mechanical revolution of rotor 12.

[0043] With N=3 pairs of coils, each Hall combination therefore covers 5A= 60°.

[0044] Figure 2 illustrates an example of such a control system for a motor 10.

[0045] A setpoint speed œ* is provided by a device external to the control system 20.

[0046] This external component can be a flight control actuator, a throttle, or any other device operable by a pilot of the aerial vehicle. This external component can also be an autopilot device, for example, for an aerial vehicle such as a drone.

[0047] This setpoint speed œ* is compared by a comparator 21 with an estimated speed œ provided by the speed estimation device 26. The result of this comparison forms a speed error (or deviation) (£ ra =œ*-œ) supplied as input to a first regulator 22

[0048] This controller can be a PI controller (PI stands for Proportional-Integral). This type of controller is commonly used in control systems to adjust a variable to follow a setpoint or target. It combines two main components: proportional (P) and integral (I), each playing a distinct role in correcting the error between the actual measurement and the setpoint.

[0049] This first regulator 22 provides a setpoint current I* to a second comparator 23, adapted to compare this setpoint current with a measured current I.

[0050] According to one embodiment, this measured intensity I can be reconstructed by a reconstruction circuit 25 from the intensities measured on the 3 pairs of coils, la, lb, le and signals from a set of position sensors 11. Knowing the position of the rotor 12 and the intensities corresponding to each phase of the applied electrical voltage, a total intensity can be deduced.

[0051] In another embodiment based on six-step motor control, the measured current value I is deduced from the phase and drive currents of the inverter.

[0052] This second comparator provides an intensity error (or deviation) (l ra =l *-l ) supplied as input to a second regulator 24. This second regulator can also be a PI (Proportional-Integral) regulator. It provides as output a setpoint for a modulator 24.

[0053] This 24 modulator is suitable for generating a set of control voltages, U a , U b , Uc, each corresponding to a group of coils of motor 10 corresponding to the same phase.

[0054] This modulator is typically a PWM (Pulse Width Modulation) modulator. This type of modulator relies on a technique to control the average power delivered to a load by modulating the duty cycle of a periodic signal. The PWM signal alternates between two states (high and low) at a fixed frequency, and the proportion of time it remains in the high state (for example) determines the average power.

[0055] This proposition is directly linked to the input setpoint provided by comparator 23. This PWM signal is also phase-shifted according to the signals from position sensors 11, so as to determine distinct voltages Ua , Ub, U c , for each group of reels.

[0056] This control system 20 is therefore based on the estimation of a speed œ of the rotor provided by the speed estimation device 26 to the speed comparator 21.

[0057] It is proposed to estimate this speed œ as an average value over a succession of position and time deviations obtained for two successive signals received from position sensors 11.

[0058] Figures 3a to 3d illustrate this calculation of an average value based on different possible sequences. The curves show the evolution of an estimated speed and an absolute speed as a function of time, represented on the x-axis.

[0059] Each figure represents a curve 31, in dashed lines, representing the absolute velocity Q, that is to say, the velocity from an absolute position sensor

[0060] Figure 3a further represents a curve 32 representative of experimental results of estimating a velocity by taking into account only the last position and time deviations (succession of 1 deviation), QMGI.

[0061] Figure 3b further represents a curve 33 representative of experimental results of estimating a velocity by averaging the last two position and time differences (succession of 2 differences), QMG2.

[0062] Figure 3c further represents a curve 34 representative of experimental results of estimating a velocity by averaging the last 4 position and time differences (succession of 4 differences), QMG4.

[0063] Figure 3d further represents a curve 35 representative of experimental results of estimating a velocity by averaging the last 6 position and time differences (succession of 6 differences), QMG6.

[0064] From these experimental studies, it was observed that: The accuracy of the velocity estimation improves when more positional and temporal variations are taken into account. As a result, curves 32 and 33 appear highly noisy. The higher the number of deviations taken into account, the longer the response time, especially for slow speeds.

[0065] It is proposed to make the number of deviations taken into account to estimate the rotor speed dependent on the latter.

[0066] It is therefore possible to substantially improve the accuracy of this estimate by increasing the number of deviations considered at high speeds (since the impact on response time is lower), and by using a smaller number of deviations at low speeds (since the impact on response time is greater). This results in a compromise that meets the accuracy and response time requirements for rotor speed estimation.

[0067] Figure 4 shows an illustrative flowchart of a proposed method that can be implemented by the rotor speed estimation device 26.

[0068] This process is iterative and relies on updating the estimated speed over time, with these updates reflecting the evolution of the rotor's actual speed. These updates are triggered by receiving signals from the set of position sensors 11 located on the motor 10. In other words, each signal transmission by the sensors (for example, Hall effect sensors) generates an event that allows the speed estimate to be updated based on the content of these signals.

[0069] In a step SI, the device 26 detects a discrete angular position 0j from these signals. In this notation, i represents the iteration of this update.

[0070] In this SI step, the time tj corresponding to the detection of an angular position 0j is also determined. This could, for example, be the time of reception of the signals provided by the Hall sensors 11. This time can be determined from a clock associated with the device 26.

[0071] In a step S2, this speed estimation device 26 determines and stores a new position deviation and a new time deviation relative to, respectively, a previous discrete angular position and a previous instant of reception of the signals transmitted by the sensors.

[0072] More precisely, in step S21, it determines a new position difference A0j between the angular position 0j detected in step SI, and a previous angular position 0j-i (detected in the previous iteration i-1): AOi = Ot — 0i- -

[0073] In an S22 step, this new position gap A0j is memorized.

[0074] This storage can be done in a stack of predefined size, for example. This predefined size can correspond to the maximum number of deviations that are taken into account for the different speed ranges (this aspect will be explained later).

[0075] Similarly, in a step S23, the device 26 determines a new time gap Atj between the instant tj of reception of the signals from the sensors 11 and a previous instant tj-i of signal reception, that is to say between two events: At L = t t —

[0076] In an S24 step, this time gap Atj is also stored, for example in a stack in the same way as the position gap A0j.

[0077] These steps S21 / S22 on the one hand and S23 / S24 on the other hand can be implemented independently of each other by the speed estimation device 26. In other words, the order in which these steps are carried out is irrelevant.

[0078] In an S3 step, the speed estimation device 26 determines a range of speeds from a set of speed ranges.

[0079] These speed ranges form a partition of possible speeds, and each is delimited by a lower and an upper threshold. The lower threshold of a range corresponds to the upper threshold of the lower range if one exists (otherwise, this threshold corresponds to zero speed). Similarly, the upper threshold of a range corresponds to the lower threshold of the upper range if one exists.

[0080] According to an illustrative example, the set of beaches includes: a lower range corresponding to an upper threshold of preferably 20 rads / s, an intermediate range (between 20 rad / s and 50 rad / s) an upper range, corresponding to a lower threshold of for example 50 rad / s.

[0081] This example has 3 ranges, but it is clear that other divisions are possible and therefore more than one intermediate range can be considered.

[0082] According to one embodiment, in order to avoid fluctuation between two speed ranges, it may be provided that the lower or upper threshold for entry into a range is slightly different from the lower or upper threshold for exiting the next range.

[0083] These ranges can be predetermined and correspond to a set of specifications.

[0084] Based on the previous estimate of the coi-i speed, the speed estimation device 26 determines the new speed range.

[0085] This determination includes a comparison between this previous speed estimate and the previous speed range.

[0086] If the previous velocity estimate œi-i falls within the previous velocity range, then the velocity range remains unchanged. Otherwise, a change of velocity range can be made. This change allows for a calculation based on a more consistent number of deviations for the new calculated velocity.

[0087] More specifically, according to embodiments: If the previous estimate of the velocity œi-i is within the previous velocity range, then the determined velocity range corresponds to the latter. If the previous velocity estimate œi-iest is lower than the lower threshold of the previous velocity range, then the determined velocity range may correspond to the immediately lower velocity range, and If the previous velocity estimate œi-iest is greater than the upper threshold of the previous velocity range, then the determined velocity range may correspond to the immediately higher velocity range.

[0088] Depending on the embodiment, the speed range is decreased or increased by a single level. This makes it possible, in particular, to regulate potential measurement errors by avoiding a situation, normally impossible in practice, of determining a speed range that is not immediately contiguous with the previous speed range.

[0089] Furthermore, according to one embodiment, in order to avoid hysteresis phenomena, the speed estimation device 26 only changes the speed range when the previous speed estimate exceeds a threshold by a predefined margin, for example around 5%.

[0090] In other words, the determination of a speed range includes a change from the previous speed range when the previous speed estimate is below or above a lower, or respectively upper, threshold of the previous speed range, reduced, or respectively increased, by a margin.

[0091] In the example given above, the step of determining a new velocity range PVj therefore depends on a previous velocity range PVj-i and the previous estimation of the velocity œi-ide in the following way: If PVj-i = PV0 (lower range): If coi-i > 21 rad / s, then PVj=PVl (intermediate range) Otherwise, PV = PV0 If PVj-i = PVl (intermediate range): If coi-i > 52.5 rad / s, then PVj=PV2 (upper range) If coi-i < 19 rad / s, then PVj = PVO (lower range). Otherwise, PV = PV1 If PVj-i = PV2 (high range): If coi-i < 47.5 rad / s, then PVj=PVl (intermediate range) Otherwise, PV = PV2

[0092] Other implementations are obviously possible. In particular, it is possible to vary the number of speed ranges, the thresholds delimiting the speed ranges and / or the margins aimed at limiting hysteresis (5% in the example above).

[0093] In step S4, the velocity estimation device 26 can estimate a velocity coi as an average value over a succession of previously stored position and time deviations. The number of these successions (the same number for both position and time deviations) is an increasing function of the velocity range.

[0094] If we denote Q i n the sum of the position differences over a succession of differences, at iteration i, and t i nthe sum of the time differences over a succession of differences, at iteration i, then the speed estimate can be written: With : >

[0095] The variable n indicates the number of gaps in the sequences of temporal gaps At k n and of position A0 k n .

[0096] This number n is determined as an increasing function of the speed range. In other words, the higher the speed range, the larger the number n.

[0097] According to one embodiment, the velocity estimate ot is zero for the lowest velocity range of the set of velocity ranges.

[0098] According to an embodiment based on 3 speed ranges, this number n of succession can be 2 for the intermediate range and 6 for the high range.

[0099] Thus, considering 3 speed ranges PVO, PV1, PV2, we can have If PV=PV0, oi = 0

[0100] This speed estimate is an update of the previous speed estimate. Each time new signals are received from the sensors 11, a new speed estimate can be determined, allowing the control system 20 to adapt the voltages U in real time a , Ub, U c applied to the coils of motor 10.

[0101] These signals are transmitted at a rate that increases with the motor speed, allowing for an estimate based on a greater number n of time deviations when the speed is high. This improves the accuracy of the estimate without impacting the reaction time due to the estimation process, thus preserving the real-time performance of the feedback loop formed by the control chain 21, 22, 23, 25. This chain therefore always allows for the determination of a setpoint for the modulator 24 according to this real-time constraint.

[0102] Of course, the present invention is not limited to the examples and embodiment described and illustrated. In particular, it is susceptible to numerous variations accessible to those skilled in the art.

Claims

DEMANDS 1. A method for estimating the speed (coi) of the rotor of a brushless motor (10), comprising updating a previous estimate (œi-i) of said speed, said update being triggered by receiving signals from a set of position sensors (11) positioned on said motor (10) and comprising steps, implemented by a speed estimation device (26), of: detection (SI) of a discrete angular position (0j) of said rotor from said signals, and determination of a time (tj) of reception of said signals; determination and storage (S2) of a new position deviation (A0j) and a new time deviation (Atj) with respect to, respectively, a previous discrete angular position (0i-i) of said rotor, and of a previous time (tj-i) of reception of said signals; determination (S3) of a range of speeds from a set of speed ranges, each delimited by a lower threshold and an upper threshold, as a function of a previous range of speeds and said previous estimate of said speed; estimation (S4) of a velocity (coi) as an average value over a succession of stored position and time deviations, the number of which is an increasing function of said velocity range.

2. A method according to claim 1, wherein said speed is estimated to be zero when the determined speed range is the lowest speed range of said assembly.

3. A method according to any one of the preceding claims, wherein said set of speed ranges comprises a lower range corresponding to an upper threshold of preferably 20 rad / s, an intermediate range and an upper range, corresponding to a lower threshold of preferably 50 rad / s.

4. A method according to the preceding claim, wherein said number of said sequences of deviations is 2 for said intermediate range, and 6 for said upper range.

5. A method according to any one of the preceding claims, wherein said determination (S3) of a speed range includes a change from said previous speed range when said previous estimate of said speed is lower or higher than a lower, respectively upper, threshold of said previous speed range, reduced, respectively increased, by a margin, said margin being preferably around 5%.

6. A method according to any one of the preceding claims, wherein said motor is a brushless DC motor.

7. A method for controlling a motor comprising a step of determining a setpoint enabling the generation of a set of control voltages for said motor, said setpoint being determined as a function of a setpoint speed and a speed estimated according to an estimation method according to one of the preceding claims.

8. A device (26) for estimating the speed (coi) of the rotor of a brushless motor (10), adapted to perform an update of a previous estimate (œi-i) of said speed, said update being triggered by the reception of signals from a set of position sensors (11) positioned on said motor (10), said speed estimation device (26) being further adapted to: detect (SI) a discrete angular position (0j) of said rotor from said signals, and determination of an instant (tj) of reception of said signals; determine and store (S2) a new position deviation (A0j) and a new time deviation (Atj) with respect to, respectively, a previous discrete angular position (0i-i) of said rotor, and a previous instant (tj-i) of reception of said signals; determine (S3) a range of speeds from a set of speed ranges, each delimited by a lower threshold and an upper threshold, as a function of a previous range of speeds and said previous estimation of said speed; estimate (S4) a speed (coi) as an average value over a succession of stored position deviations and time deviations, the number of which is an increasing function of said range of speeds.

9. Control system (20) of a motor (10) comprising a device (26) for estimating a speed of the rotor of said motor according to the preceding claim, and a control chain (21, 22, 23, 25) adapted to determine a setpoint to a modulator (24) adapted to generate a set of control voltages for said motor (10), said setpoint being determined as a function of a setpoint speed and said speed estimate.

10. Aerial vehicle comprising at least one control system according to the preceding claim, said set speed being provided by a flight control actuator.

Images