Method for determining the position and / or speed of a rotor of an electrical machine by processing the signals of a position sensor

Description

Technical field

[0001] The present invention relates to the field of determining the position and / or speed of a rotor of an electrical machine.

[0002] Precise knowledge of the angular position and / or angular velocity of an electrical machine rotor is useful in particular for the control and monitoring of electrical machines. Indeed, for the control of such electrical machines, in particular for the control of torque and / or rotational speed, the position and / or speed of the rotor are often essential data.

[0003] There are a wide variety of position sensors that have been developed for this application, including magnetostrictive sensors, encoders, resolvers, GMR sensors (based on the Giant Magnetoresistance Effect), and inductive sensors. Such sensors can generate two signals, a first cosine signal and a second sinusoidal signal, from which the position and / or speed of the rotor can be reconstructed.

[0004] Ideally, these signals should have the same amplitude, no offset, and should be orthogonal (with a phase shift strictly equal to 90°). However, these signals may include harmonics, which, when reconstructing the rotor position and / or speed, can generate significant inaccuracies for determining the rotor position or speed. In addition, there may also be a phase shift between these two signals (related, for example, to the position of the position sensor), which can also generate an error for determining the rotor position or speed. These inaccuracies and errors have a significant impact on the control of the electrical machine or on the monitoring of the electrical machine. Indeed, an error in the rotor position can, for example, generate a torque setpoint that is not adapted to the use of the electrical machine, or can generate torque harmonics, etc. Previous technique

[0005] In order to limit the impact of harmonics and phase shift, several technical solutions have been developed.

[0006] For example, patent application DE 102014226604 describes a method for correcting the angular position of the rotor. This method uses the measurement of two position sensors located at 90° to each other around the rotor, and estimates a correction value that is a function of the phase difference of the signals from the two position sensors. However, this method only corrects the phase shift and not the harmonics.

[0007] Patent application DE 102016220188 describes a method for correcting measurements from a rotation sensor generating sine and cosine signals, using static and dynamic correction. The static correction is based on tables, and the dynamic correction is intended to take into account the difference in amplitude between the signals and the offset differences. However, this method does not allow for continuous correction of harmonics. Furthermore, the use of tables may not provide optimal correction and requires the preparation of complex tables, which may depend, among other things, on the rotation speed and the load of the electrical machine.

[0008] US patent application 2019 / 0031046 relates to the elimination of offset signals from position measurements of an electrical machine. This method is based on the addition and subtraction of signals and on the use of integrators by means of a motion-state filter. This method therefore only eliminates offsets, but does not determine and eliminate harmonics.

[0009] Patent application EP2387145 describes a method of removing disturbances for controlling a variable speed motor, in which a Fourier reconstruction and a phase-locked loop are applied without phase shift correction.

[0010] Patent application US2017 / 284826 describes a method for interpolating position measurements from an encoder, without a phase-locked loop. Summary of the invention

[0011] The present invention aims to determine the speed and position of a rotor, accurately, without error. The invention relates to a method for determining the angular speed and / or angular position of a rotor of an electrical machine, by means of a determination of the harmonics, which is implemented by means of a closed loop which comprises a harmonic observer, a phase shift correction, and a first phase-locked loop which estimates the position of the rotor.

[0012] Furthermore, the invention relates to a method for determining the speed and / or position of an electrical machine rotor by accurately determining the harmonics, by continuously correcting the measurement signals only when the harmonic observer is convergent, the corrected signals being used to determine the position and / or speed of the rotor. The determination of the harmonics is implemented by means of a closed loop which comprises a harmonic observer, a phase shift correction, and a first phase-locked loop which estimates the position of the rotor.

[0013] The invention relates to a method for determining the position and / or speed of a rotor of an electrical machine by means of a position sensor of said rotor, said position sensor generating a cosine signal and a sinusoidal signal.For this method, a closed loop is implemented which comprises a harmonic state observer, a correction of said phase shift of said cosine and sinusoidal generated signals, and a first PLL phase-locked loop, said harmonic state observer connecting said cosine and sinusoidal generated signals and a value of said rotor position estimated by said first PLL phase-locked loop to said harmonics, said correction of said phase shift identifying and correcting the phase shift of said cosine and sinusoidal generated signals by means of said harmonics determined by said harmonic state observer, and said first PLL phase-locked loop estimating the position and / or speed of said rotor from said corrected cosine and sinusoidal signals.

[0014] According to one embodiment, the following steps are implemented: a) The harmonics of said cosine and sinusoidal generated signals are determined by means of said closed loop which comprises said state observer of said harmonics, said correction of said phase shift of said cosine and sinusoidal generated signals, and said first PLL phase-locked loop; b) It is determined whether said state observer of said harmonics is convergent; c) It is continuously corrected said cosine and sinusoidal generated signals by updating said determined harmonics when said state observer of said harmonics is convergent; and d) It is determined said position and / or said speed of said rotor by means of a second PLL phase-locked loop from said corrected cosine and sinusoidal signals.

[0015] Advantageously, it is determined whether said state observer of said harmonics is convergent by verifying the equation: y a − y ^ a 2 + y b − y ^ b 2 < ε with y a and yb said generated cosine and sinusoidal signals, ŷ a And ŷ b the signals of said signals reconstructed by means of said estimated position and e a predetermined threshold.

[0016] Preferably, said first and second phase-locked loops PLL comprise a proportional-integral regulator, and an integrator.

[0017] Advantageously, the transfer function of said first and second PLL phase-locked loops is written: θ ^ s θ s = 1 + K p K i ∗ s 1 + K p K i ∗ s + 1 K i ∗ s 2 with θ the position of said rotor, θ̂ the estimated position of said rotor, s the Laplace parameter, K p the proportional coefficient of said proportional integral regulator, K i the coefficient of said proportional integral regulator.

[0018] According to one implementation, said integral coefficient K i of said first phase-locked loop is less than said integral coefficient K i of said second phase-locked loop.

[0019] According to one aspect, said cosine and sinusoidal signals are corrected by filtering said determined harmonics, and possibly by correcting the phase shift of said cosine and sinusoidal signals.

[0020] According to one characteristic, said state observer of said harmonics implements a transfer function: y ^ k y = α ∗ s s 2 + k ∗ ω 2 1 + ∑ k = 0 → n α ∗ s s 2 + k ∗ ω 2 with s the Laplace parameter, α a gain, k the order of the harmonic considered, n the number of harmonics of said cosine and sinusoidal signals, w the fundamental frequency, y the generated signal considered among the cosine and sinusoidal signals, and ŷ k the estimated harmonic of order k of said generated signal considered among the cosine and sinusoidal signals.

[0021] According to one embodiment, said gain α is lower than said fundamental frequency, preferably said gain α is lower than one tenth of said fundamental frequency ω.

[0022] According to an embodiment option, said state observer of said harmonics further determines said fundamental coefficients of said harmonics and said offsets of said generated cosine and sinusoidal signals.

[0023] According to one implementation, said phase shift between said cosine and sinusoidal generated signals is identified by means of said fundamental coefficients of said harmonics by means of an arctangent function.

[0024] Advantageously, said phase shift Φ between said cosine and sinusoidal generated signals is determined by means of the equation: ϕ = atan L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ with L 1 b * sin( ϕ ) = -L̂ 11 b * cos( Dth ) - L̂ 12 b * sin( Dth) L 1 b * cos( ϕ ) = L̂ 12 b * cos( D i ) - L̂ 11 b * sin( Dth ), and with L 1b the fundamental coefficient of one of said cosine and sinusoidal generated signals, L̂ 11 b And L̂ 12 b the fundamental coefficients determined by said state observer of said harmonics, Δθ the difference between the position of said measured rotor and the position of said rotor estimated by said first phase-locked loop PLL.

[0025] According to one embodiment, said phase shift of one of said generated cosine and sinusoidal signals is corrected by means of the equation y fb − correction = L 1 b ∗ y fb − y fa L 1 a ∗ L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ with y fb-correction the fundamental of said generated signal corrected by one of said cosine and sinusoidal generated signals, L 1a and L 1b the fundamental coefficients of said cosine and sinusoidal generated signals, y fa and y fb the fundamentals of said measured cosine and sinusoidal generated signals, and Φ said phase shift.

[0026] In one aspect, said position sensor is a magnetostrictive sensor, an inductive sensor, an encoder, a GMR sensor, an AMR sensor, a TMR sensor or a resolver.

[0027] Furthermore, the invention relates to a method for controlling an electrical machine, said electrical machine comprising a position sensor of the rotor of said electrical machine, said position sensor generating a cosine signal and a sinusoidal signal, in which the following steps are implemented: a) said position and / or said speed of said rotor is determined by means of said method according to one of the preceding characteristics and said signals generated by said position sensor; and b) said electrical machine is controlled as a function of said predetermined position and / or speed.

[0028] Other characteristics and advantages of the method according to the invention will appear on reading the following description of non-limiting examples of embodiments, with reference to the figures appended and described below. List of figures

[0029] There figure 1 illustrates the steps of the method according to a first embodiment of the invention. The figure 2 illustrates the steps of the method according to a second embodiment of the invention. The figure 3 illustrates the construction of a phase-locked loop according to one embodiment of the invention. The figure 4illustrates the reference rotation speed for an example of application of the method according to the invention. The figure 5 illustrates the rotation speed determined by the method according to an embodiment of the invention for the example of application of the figure 4 . There figure 6 illustrates the position error between the reference position and the position determined by the method according to one embodiment for the application example of the figure 4 . Description of the modes of realization

[0030] The present invention relates to determining the position of a rotor of an electrical machine. The electrical machine is equipped with an angular position sensor which, during measurement, generates a cosine signal and a sinusoidal signal. In the following description, these two signals are called measurement signals. Furthermore, in the following description, the angular position of the rotor is called position, and the angular speed of the rotor is called speed. The present invention is suitable for any electrical machine for which the position is measured. The present invention is particularly suitable for permanent magnet-assisted synchro-reluctant machines, which are very sensitive to position measurement.

[0031] The position sensor can be selected from magnetostrictive sensors, inductive sensors, encoders and resolvers, a GMR sensor (based on Giant Magnetoresistance), an AMR sensor (based on Anisotropic Magnetoresistance), a TMR sensor (based on Tunnel Effect Magnetoresistance) or any other sensor capable of generating a cosine measurement signal and a sinusoidal measurement signal.

[0032] A harmonic is a primary decomposition element of a periodic function, here the periodic function corresponds to the measurement signals. The fundamental is the harmonic of order one of the measurement signals, i.e. the first harmonic. The continuous shift of the measurement signals is called offset.

[0033] Preferably, the method according to the invention can implement a single position sensor, thus limiting the inaccuracies linked to the multiplication of sensors, and to the difficulty of precisely positioning several sensors.

[0034] According to a first embodiment, the method can implement a closed loop making it possible to determine the harmonics of the measurement signals and to deduce therefrom the position and / or the speed of the rotor. The closed loop will be described in detail in relation to step 1) in the remainder of the description. The steps of the closed loop can be implemented by a computer or electronic system, in particular a computer or electronic system which controls the electrical machine.

[0035] There figure 1illustrates, schematically and in a non-limiting manner, the steps of the method according to the first embodiment of the invention. First, the position sensor CAP generates the measurement signals y. Then, the harmonics of the measurements are determined by means of a closed loop BF, which comprises a harmonic state observer OBS, a phase shift corrector CORΦ and a first phase-locked loop PLL1. The harmonic state observer OBS determines the harmonics from the measurement signals y and the rotor position estimated θ obs by the first phase-locked loop PLL1. The harmonic state observer OBS also determines the signals y_hf, formed by the suppression of harmonics of order strictly higher than 1 from the measurement signals y.The phase shift corrector CORΦ identifies and corrects the phase shift as well as the offset and amplitudes of the measurement signals y_hf to form a corrected signal yd from the identified harmonics. The first phase-locked loop PLL1 estimates an estimated rotor position θ obs from the corrected signal yd. For this first embodiment, the estimated rotor position and / or speed are obtained by the first phase-locked loop PLL1.

[0036] The method according to the second embodiment of the invention comprises the following steps: 1) Determination of harmonics 2) Determination of the convergence of the harmonic observer 3) Correction of the observed harmonics 4) Correction of the measurement signals 5) Determination of the position and / or speed of the rotor

[0037] These steps will be detailed in the remainder of the description. These steps may be implemented by a computer or electronic system, including a computer or electronic system that controls the electrical machine.

[0038] There figure 2illustrates, schematically and in a non-limiting manner, the steps of the methods according to the second embodiment of the invention. Firstly, the position sensor CAP generates the measurement signals y. Then, the harmonics of the measurements are determined by means of a closed loop BF, which comprises a harmonic state observer OBS, a phase shift corrector CORΦ and a first phase-locked loop PLL1. The harmonic state observer OBS determines the harmonics H from the measurement signals y and the rotor position estimated θ obs by the first phase-locked loop PLL1. The harmonic state observer OBS also determines the signals y_hf, formed by the suppression of harmonics of order strictly higher than 1 from the measurement signals y.The phase shift corrector CORΦ identifies and corrects the phase shift as well as the offset and the amplitude of the measurement signals y_hf to form a corrected signal yd from the identified harmonics H. The first phase-locked loop PLL1 estimates an estimated rotor position θ obs from the corrected signal yd. Furthermore, the method according to the invention comprises a step of correcting the harmonics COROBS if the convergence C of the harmonic state observer OBS is ensured. The corrected harmonics are denoted Hc. Furthermore, the method according to the invention relates to a step of removing the harmonics CORH from the measurement signal y from the determined corrected harmonics Hc. The corrected signal is denoted yc. Furthermore, the method according to the invention comprises a second phase-locked loop PLL2 which estimates the position . θ̂ of the rotor and / or the rotor speed ohfrom the signal yc. According to one embodiment of the invention, the method may further comprise feedback of the estimated position θ̂ by the second phase-locked loop PLL2 for the correction of harmonics CORH. For this second embodiment, the estimated rotor position and / or speed are obtained by the second phase-locked loop PLL2. 1. Determination of harmonics

[0039] In this step, the harmonics of the position sensor measurement signals are determined using a closed loop. A closed loop is a sequence of steps in which one of the step outputs also serves as an input for at least one other step preceding it. In other words, the closed loop includes at least one feedback.

[0040] The closed loop includes: A harmonic state observer, which connects the measurement signals and the estimated rotor position to the harmonics of the measurement signals, and filters harmonics of order strictly greater than 1 from the measurement signals, A phase shift corrector, which identifies and corrects the amplitude, offset and phase of the filtered measurement signals using harmonics of order less than or equal to 1, determined by the harmonic state observer, A first phase-locked loop, which estimates the rotor position from the corrected signals, and A feedback of the position estimated by the first phase-locked loop to an input of the harmonic state observer.

[0041] A state observer is an extension of a model represented as a state representation. When the state of a system is not measurable, an observer is designed to reconstruct the state from a model of the dynamic system and measurements of other quantities.

[0042] A phase-locked loop (PLL) is an electronic or computer circuit that allows the output phase or frequency of a system to be controlled by the phase or frequency of the input signal. It can also control an output frequency by a multiple of the input frequency. The first phase-locked loop allows for better identification of harmonics.

[0043] According to one embodiment of the invention, the harmonic state observer can implement, for each harmonic of order k, a transfer function of the form: y ^ k y = α ∗ s s 2 + k ∗ ω 2 1 + ∑ k = 0 → n α ∗ s s 2 + k ∗ ω 2 with s the Laplace parameter, α a gain of the state observer, k the order of the harmonic considered, n the number of harmonics of the measurement signals, w the fundamental frequency with i = oh * t, θ being the angular position, y the generated signal considered among the cosine and sinusoidal signals, and ŷ k the estimated harmonic of order k of the generated signal considered among the cosine and sinusoidal signals.

[0044] Advantageously, the gain of the state observer α can be determined so as to be less than the fundamental frequency w. Preferably, the gain of the state observer α can be less than one tenth of the fundamental frequency w. The gain α makes it possible to adjust the convergence of the harmonic state observer. According to an exemplary embodiment, the gain α can be 10. This value allows rapid convergence without significant distortion of the observed signal.

[0045] According to one embodiment, the transfer function can be determined by means of the following operations: The measurement signals, denoted respectively y a and yb , which include harmonics, can be written:

[0046] Or again:

[0047] In these equations and in the following description, the index a denotes the first measurement signal (e.g. the cosine signal), and the index b denotes the second measurement signal (e.g. the sinusoidal signal). In these equations, i = ω × twith w the fundamental frequency, L 11 a, L 12a , L 11b , L 12b the fundamental coefficients of the measurement signals (i.e. the coefficients linked to the fundamental of the harmonics), L i1a , L i2a , L i1b , L i2b (with i varying between 2 and n, n being the highest frequency harmonic considered) the coefficients of the harmonics of the measurement signals, L 0a and L 0b the offsets of the measurement signals (i.e. the continuous shifts).

[0048] To simplify the writing, we can note the vectors L and W (without the indices a and b): L = L 0 L 11 L 12 L 21 L 22 ⋯ L n 1 L n 2 T , W = 1 sin θ cos θ sin 2 θ cos 2 θ ⋯ sin nθ cos nθ T

[0049] And we can define the vector of the estimation of the coefficients L̂ in the following manner: L ^ = L 0 ^ L 11 ^ L 12 ^ L 21 ^ L 22 ^ ⋯ L n 1 ^ L n 2 ^ T

[0050] We can then write: y = L T ∗ W , et y ^ = L ^ T ∗ W

[0051] According to the following documents: Osowski, S. (1992). Neural network for estimation of harmonic components in a power system. IEEE Generation, Transmission and Distribution, (pp. 129-135). and Siyu Leng, WL-Y. (2009). Active Power Filter for Three-Phase Current. IEEE American Control Conference, (pp. 2140-2147). MO, USA.

[0052] We can write: L ^ = P ∗ y − y ^ ∗ W

[0053] Where P is the diagonal matrix of dimension (2n+1)*(2n+1) which is defined by: P = diag α

[0054] With α the gain of the state observer.

[0055] The coefficient estimator L̂ can then be written as follows: L 0 ^ . = α ∗ y − y ^ L 11 ^ . = α ∗ y − y ^ ∗ sin θ L 12 ^ . = α ∗ y − y ^ ∗ cos θ L 21 ^ . = α ∗ y − y ^ ∗ sin 2 θ L 22 ^ . = α ∗ y − y ^ ∗ cos 2 θ ⋯ L n 1 ^ . = α ∗ y − y ^ ∗ sin n θ L n 2 ^ . = α ∗ y − y ^ ∗ cos n θ

[0056] From these equations, we can extract the estimate of the fundamental part noted ŷ f by removing the noted harmonics h and the offsets noted offset : y ^ f = y − y ^ h − y ^ offset Or y ^ offset = L 0 ^ et y ^ h = ∑ i = 2 n L ι 1 ^ sin iθ + L ι 2 ^ cos iθ

[0057] Using the Laplace transform to form the following equation: y t − y ^ t = y t − ∑ k = 0 n y ^ k t où y k t = L ^ k 1 sin k ∗ θ + L ^ k 2 cos k ∗ θ

[0058] With k the order number of the harmonic. We can then write: y s − y ^ s = y s − ∑ k = 0 → n y ^ k s

[0059] With y ^ k s = L L ^ k 1 sin k ∗ θ + L L ^ k 2 cos k ∗ θ

[0060] With the Laplace transform.

[0061] We can write: y ^ k s = y s − y ^ s ∗ α ∗ s s 2 + k ∗ ω 2

[0062] The transfer function can then be written for each harmonic of order k: y ^ k y = α ∗ s s 2 + k ∗ ω 2 1 + ∑ k = 0 → n α ∗ s s 2 + k ∗ ω 2

[0063] For the fundamental (k=1), the transfer function between the fundamental ŷ f and the measure y can be written: H s = y ^ f s y s = 1 + α ∗ s s 2 + ω 2 1 + ∑ k = 0 → n α ∗ s s 2 + k ∗ ω 2

[0064] According to an implementation of the invention, the harmonic state observer can further determine the fundamental coefficients of said harmonics and the offsets of the measurement signals.

[0065] According to one embodiment, for the phase shift correction step, the phase shift between the measurement signals, preferably between the fundamentals of the measurement signals, can be identified by means of the fundamental coefficients of the harmonics. This identification of the phase shift can be implemented by means of an arctangent function.

[0066] Advantageously, the phase shift Φ between the fundamentals of the measurement signals can be determined using the equations: ϕ = atan L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ

[0067] By using L 1 b ∗ sin ϕ = − L ^ 11 b ∗ cos Δθ − L ^ 12 b ∗ sin Δθ L 1 b ∗ cos ϕ = L ^ 12 b ∗ cos Δθ − L ^ 11 b ∗ sin Δθ . with L 1b the fundamental coefficient of one of said cosine and sinusoidal generated signals, L̂ 11 b And L̂ 12 b the fundamental coefficients determined by said harmonic state observer, Δθ the difference between the measured rotor position and the rotor position estimated by said first phase-locked loop PLL.

[0068] Indeed, for this step, we can implement the following operations: With the hypothesis: ŷ fa = yes And ŷ fb = yb , (the index f indicating the fundamental), we can write the following equations: L ^ 11 a sin θ ^ + L ^ 12 a cos θ ^ = L 1 a ∗ sin θ . L ^ 11 b sin θ ^ + L ^ 12 b cos θ ^ = L 1 b ∗ cos θ + ϕ

[0069] With θ̂ the estimated value of the rotor position (which corresponds to θ obs on the figure 1 or on the figure 2 ). The fundamental signal coefficients L a and L b can be calculated by the following equations: L 1 a = L ^ 11 a 2 + L ^ 12 a 2 , L 1 b = L ^ 11 b 2 + L ^ 12 b 2

[0070] By posing i = θ̂ + Δ i, we can write: L 1 a ∗ sin θ = L 1 a ∗ sin θ ^ + Δ θ = L 1 a ∗ sin θ ^ ∗ cos Δθ + L 1 a ∗ cos θ ^ ∗ sin Δθ , And L 1 b ∗ cos θ + ϕ = L 1 b ∗ cos θ ^ + Δ θ + ϕ = = L 1 b ∗ cos θ ^ ∗ cos θ ^ ∗ cos Δ θ + ϕ − L 1 b ∗ sin θ ^ ∗ sin Δ θ + ϕ

[0071] We can then obtain: L ^ 11 a = L 1 a ∗ cos Δθ , L ^ 12 a = L 1 a ∗ sin Δθ L ^ 11 b = − L 1 b ∗ sin Δ θ + ϕ , L ^ 12 b = L 1 b ∗ cos Δ θ + ϕ

[0072] From these equations, we can show that: cos Δθ = L ^ 11 a L 1 a , sin Δθ = L ^ 12 a L 1 a L 1 b ∗ sin ϕ = − L ^ 11 b ∗ cos Δθ − L ^ 12 b ∗ sin Δθ L 1 b ∗ cos ϕ = L ^ 12 b ∗ cos Δθ − L ^ 11 b ∗ sin Δθ

[0073] By injecting the last two equations, we can identify the phase shift Φ in the following equation: ϕ = atan L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ

[0074] From the identified phase shift, one of the two measurement signals can be corrected to eliminate the phase shift. To do this, a signal without phase shift can be reconstructed using the determined values of the fundamentals of the harmonics of the measurement signals, the values of the fundamental coefficients, and the identified phase shift angle.

[0075] According to one embodiment of the invention, the phase shift of one of said measurement signals is corrected by means of the equation y fb − correction = L 1 b ∗ y fb − y fa L 1 a ∗ L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ with y fb-correction the fundamental of the measurement signal considered, L 1a and L 1b the fundamental coefficients of the measurement signals, y fa and y fb the fundamentals of the measurement signals and Φ said phase shift.

[0076] Indeed, by definition: y fa = L 1 a ∗ sin θ , y fb = L 1 b ∗ cos θ + ϕ

[0077] And we want to establish the corrected fundamental signal defined by: y fb − correction = L 1 b ∗ cos θ

[0078] If we use the following development: cos θ = cos θ + ϕ − sin θ ∗ sin ϕ cos ϕ

[0079] We can obtain: y fb − correction = L 1 b ∗ cos θ + ϕ − L 1 b ∗ sin θ ∗ sin ϕ cos ϕ

[0080] Replacing with the definition of fundamental signals, we can have: y fb − correction = L 1 b ∗ y fb − y fa L 1 a ∗ L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ

[0081] According to one embodiment of the invention, the first phase-locked loop may notably comprise a proportional-integral regulator and an integrator.

[0082] Preferably, the first phase-locked loop may be in accordance with the embodiment of the figure 3 . For this embodiment, the first phase-locked loop comprises as input a sinusoidal signal sin(θ) and a cosine signal -cos(θ). The sinusoidal signal sin(θ) is multiplied by a signal corresponding to the cosine of the estimated angular position θ̂ (for this step θ̂ corresponds to I see of the figures 1 and 2) by the output of the first phase-locked loop. The cosine signal -cos(θ) is multiplied by a signal corresponding to the sine of the estimated angular position θ̂ by the output of the first phase-locked loop. The outputs of the multipliers are then added together. This produces a signal e which corresponds to sin ( i ) × cos ( θ̂ ) - cos ( i ) × sin ( θ̂ ) which is equal to sin ( θ - θ̂ ) . The signal e then passes into a PI proportional-integral regulator to estimate the rotor rotation speed oh The PI regulator includes a proportional coefficient denoted K p , and an integrator coefficient K i . This signal is then integrated into an integrator, which corresponds to the function 1 s in the Laplace domain, to estimate the rotor position θ̂ .

[0083] For this embodiment, the transfer function of the first phase-locked loop can be written: θ ^ s θ s = 1 + K p K i ∗ s 1 + K p K i ∗ s + 1 K i ∗ s 2 with θ the position of the rotor, θ̂ the estimated position of the rotor by the first phase-locked loop, s the Laplace parameter, K p the proportional coefficient of the proportional-integral regulator, K i the coefficient of the proportional-integral regulator.

[0084] Indeed, the first phase-locked loop allows us to write the following relation: θ ^ s = 1 s ∗ K p + K i s ∗ ϵ s

[0085] By means of a linear approximation such that ϵ = sin θ − θ ^ ≈ θ − θ ^ we can obtain the following equation: θ ^ s = 1 s ∗ K p + K i s ∗ θ s − θ ^ s

[0086] Which allows us to obtain the transfer function described above.

[0087] According to one aspect of the invention, the coefficients K p and K i verify the following inequalities K p >0 and K i >0 so that the phase-locked loop is stable.

[0088] The natural frequency of the phase-locked loop oh PLL is worth ω PLL = K i .

[0089] For good performance of the phase-locked loop, the coefficients K p and K i are determined so that the defined quality factor m PLL , as defined below, is close to 1. m PLL = ω PLL 2 ∗ K p K i

[0090] For the first embodiment, the estimated rotor position and / or speed are obtained by the first phase-locked loop.

[0091] For the second embodiment, steps 2) to 5) described below are implemented. 2. Determination of the convergence of the harmonic observer

[0092] In this step, the convergence of the harmonic state observer of the closed loop is determined, so that the harmonic correction can be activated only in the case where the harmonic state observer is convergent. Thus, only good harmonic estimates are kept and corrections based on non-converged values are avoided, which avoids inaccuracies and errors in determining the position and / or speed of the rotor. Thus, the determination of the position and / or rotational speed of the rotor is improved.

[0093] The convergence of the harmonic state observer can be determined by any means.

[0094] According to one embodiment of the invention, the convergence of the harmonic state observer can be determined by means of the following equation: y a − y ^ a 2 + y b − y ^ b 2 < ε

[0095] With e a predetermined threshold, yhas and yb the measurement signals and ŷ a And ŷ b the estimated signals reconstructed from the harmonics determined by the harmonic state observer and the position determined by the first phase-locked loop.

[0096] According to one embodiment of the invention, the convergence of the observer can be determined when the above equation is verified over a constant or variable time interval, for example over one or more electrical towers. 3. Harmonic correction observed

[0097] In this step, the observed harmonics are corrected using the first phase-locked loop, in order to synchronize them with the second phase-locked loop. In the general case, the electrical positions of the two phase-locked loops are different by a quantity denoted Δ θ PLL . The reference position, noted here i, is estimated by the second phase-locked loop, which has faster kinematics than the first phase-locked loop. The observer is considered to be convergent. We then have, for example for the sinusoidal signal noted a: y ^ ka = y ka L ^ k 1 a PLL 1 sin k θ PLL 1 ^ + L ^ k 2 a PLL 1 cos k θ PLL 1 ^ = L ka ∗ sin kθ + ϕ ka

[0098] Either : L ^ k 1 a PLL 1 = L ka ∗ cos k Δ θ PLL + ϕ ak et L ^ k 2 a PLL 1 = L ka ∗ sin kΔθ PLL + ϕ ak

[0099] We seek the values of the harmonics associated with the second phase-locked loop according to: L ^ k 1 a PLL 2 = L ka ∗ cos ϕ ak et L ^ k 2 a PLL 2 = L ka ∗ sin ϕ ak

[0100] We obtain by calculation: L ^ k 1 a PLL 2 = L ^ k 1 a PLL 1 cos kΔθ PLL + L ^ k 2 a PLL 1 sin kΔθ PLL L ^ k 2 a PLL 2 = L ^ k 2 a PLL 1 cos kΔθ PLL − L ^ k 1 a PLL 1 sin kΔθ PLL

[0101] It can be noted that the above calculation consists, for each harmonic, of a compensatory rotation of the observed harmonics, of angle equal to the angle difference observed by the two phase-locked loops multiplied by the order of the harmonic considered.

[0102] According to one embodiment of the invention, the determination of the quantity PLLcan be done as described below. We recall that we have set by definition ( y 1 a is the phase reference of the system): y 1 a = L 11 a ∗ sin θ

[0103] SO L 12 a is equal to 0 by definition as well as ϕ a 1 .

[0104] Taking up the equation L̂ k2a PLL 1 = L ka * sin( kΔθ PLL + ϕ ak ) , applied to harmonic 1, we find: L̂ 12 aPLL 1 = L ka * sin( PLL ) , and similarly: L̂ 11 PLL 1 = L ka * cos( PLL ), from which we can calculate L ka and also PLL . 4. Correction of measurement signals

[0105] In this step, the measurement signals are continuously corrected so as to eliminate the effects of the harmonics determined in step 1) and corrected in step 3), only when the harmonic state observer has been determined to be convergent in step 2). In other words, the correction of the measurement signals is activated when the harmonic state observer gives reliable observations.

[0106] When the harmonic state observer is not convergent, the measurement signals are corrected using the last observed harmonic values when the observer was convergent, and the signals are used in step 5).

[0107] According to one embodiment of the invention, the measurement signals can be continuously corrected by filtering the harmonics determined in steps 1) and 3).

[0108] The filtering of harmonics in the measurement signals is obtained according to y c = y − L 0 a − L 21 a sin 2 θ − L 22 a cos 2 θ − L 31 a sin 3 θ − L 32 a cos 3 θ − ⋯ − L n 1 a sin nθ − L n 2 a cos nθ

[0109] According to one embodiment of the invention, the phase shift, amplitude and offset of the signals can then be corrected. yc , by a method identical to that used in step 1) 5. Determination of the position and / or speed of the rotor

[0110] In this step, the position and / or speed of the rotor is determined by means of a second phase-locked loop from the corrected measurement signals when the harmonic state observer is convergent (at the output of step 3), and from the uncorrected measurement signals when the harmonic state observer is divergent.

[0111] The second phase-locked loop may have the same structure as the first phase-locked loop (according to the figure 3). The building blocks and transfer function are identical to those described in step 1). However, the coefficients of the PI controller may be different. In particular, the integral coefficient K i of the first phase-locked loop may be smaller than the integral coefficient K i of the second phase-locked loop. The integral coefficient K i of the first phase-locked loop may be related to the resonance of the mechanical system, while the integral coefficient K i of the second phase-locked loop may depend on the electrical bandwidth. In this way, the first phase-locked loop has slower kinematics than the second phase-locked loop.

[0112] Furthermore, the invention relates to a method for controlling an electric machine, the electric machine comprising an angular position sensor of the rotor of the electric machine. The position sensor generates a cosine signal and a sinusoidal signal. The control method comprises the following steps: a) The position and / or speed of the rotor is determined by means of the method for determining the position and / or speed of the rotor according to any of the variants or combinations of variants described above, the method being applied to the measurement signals of the position sensor; and b) The electrical machine is controlled as a function of the determined position and / or speed.

[0113] According to one embodiment of the invention, it is possible to control in particular the torque, or the rotation speed of the electric machine.

[0114] The invention also relates to a method for monitoring an electrical machine, the electrical machine comprising an angular position sensor of the rotor of the electrical machine. The position sensor generates a cosine signal and a sinusoidal signal. The monitoring method comprises the following steps: a) The position and / or speed of the rotor is determined by means of the method for determining the position and / or speed of the rotor according to any of the variants or combinations of variants described above, the method being applied to the measuring signals of the position sensor; and b) The electrical machine is monitored as a function of the determined position and / or speed.

[0115] According to one embodiment of the invention, the monitoring method may comprise a step of diagnosing abnormal operation of the electrical machine. In this case, the monitoring method may comprise a step of controlling the electrical machine in order to take into account the abnormal operation of the electrical machine. For example, the control in the event of abnormal operation of the electrical machine may consist of stopping the electrical machine.

[0116] It goes without saying that the invention is not limited to the embodiments of the steps described above as examples; on the contrary, it encompasses all variant embodiments. Examples

[0117] The characteristics and advantages of the method according to the invention will appear more clearly on reading the application example below.

[0118] For this example, the method for determining the position and speed of the rotor according to the second embodiment of the invention is simulated. For this simulation, theoretical measurement signals of the form are considered: y a = 0.95 ∗ sin ωt + 0.05 + 0.09 ∗ sin 2 ωt − 0.04 ∗ cos 2 ωt + 0.02 ∗ sin 3 ωt + w a ,

[0119] For this example, the fundamental phase shift Φ is 3°, the noises are denoted wa and wb, and the gain of the harmonic state observer α is 10.

[0120] There figure 4illustrates the evolution of the reference rotation speed ω ref in rad / s as a function of time T in s for the simulation. The rotation speed ω ref considered comprises five phases: a first phase between 0 and 3s with a small increase in the rotation speed ω ref , a second phase between 3 and 4s with a strong increase in the rotation speed ω ref , a third phase between 4 and 5.5 s of stability of the rotation speed ω ref , a fourth phase between 5.5 and 7s of strong decrease in the rotation speed ω ref , and a fifth phase between 7 and 10 s of stability of the rotation speed ω ref .

[0121] For the simulation, we use for the two phase-locked loops according to the structure illustrated in figure 3. For the first phase-locked loop, a natural frequency ω PLL of 60 rad / s is chosen, which implies an integral coefficient K i of 3600. For the second phase-locked loop, a natural frequency ω PLL of 400 rad / s is chosen, which implies an integral coefficient K i of 160000. For the first phase-locked loop, the proportional coefficient K p is 120. For the second phase-locked loop, the proportional coefficient K p is 800. Thus, for each phase-locked loop, the quality factor m PLL is equal to 1.

[0122] For this example, the figure 5 illustrates the estimated rotation speed ω in rad / s by the method according to the invention as a function of time T in s, at the output of the second phase-locked loop. The figure 6illustrates the difference Δθ in rad between the angle estimated by the method according to the invention and the reference angle as a function of time T in s, at the output of the second phase-locked loop. Initially, we observe that the estimated speed is similar to the reference speed. We also observe that in the low-speed zone (up to approximately 85 rad / s), the harmonic observer is not yet activated, so there is a large estimation error. However, as soon as the rotation speed reaches the threshold of 85 rad / s (this threshold corresponds substantially to 2 times the natural frequency of the first phase-locked loop), the harmonic observer works well, which implies the convergence of the harmonic observer, which creates good estimates of speed and position by the second phase-locked loop. In particular, from the second phase to the fifth phase, the speed and position estimates remain reliable. It can also be noticed that when the rotational speed decreases (fourth phase), the estimation of speed and position still remains good even if the rotational speed is lower than 2 times the natural frequency of the second phase-locked loop.

[0123] Therefore, the method according to the invention makes it possible to precisely and reliably determine the position and rotational speed of a rotor of an electrical machine.

Claims

1. Method for determining the position and / or speed of a rotor of an electric machine by means of a position sensor (CAP) for sensing a position of said rotor, said position sensor (CAP) generating a cosinusoidal signal and a sinusoidal signal, implementing - a closed loop (BF) comprising a harmonics state observer (OBS), - a correction (CORΦ) of a phase shift of said cosinusoidal and sinusoidal generated signals, - a first phase-locked loop (PLL1), - said harmonics state observer (OBS) relating said cosinusoidal and sinusoidal generated signals and a value of said rotor position (θobs) estimated by said first phase-locked loop (PLL1) with said harmonics, - said correction (CORΦ) of a phase shift identifying and correcting the phase shift of said cosinusoidal and sinusoidal generated signals by means of said harmonics determined by said harmonics state observer (OBS), - said first phase-locked loop (PLL1) estimating the position and / or speed of said rotor based on said corrected cosinusoidal and sinusoidal signals, characterized in that the determining method implements a second phase-locked loop (PPL2), - said second phase-locked loop (PLL2) determining said position and / or speed of said rotor by carrying out the following steps: a) the harmonics of said cosinusoidal and sinusoidal generated signals are determined by means of said harmonics state observer (OBS) of said closed loop (BF) comprising said harmonics state observer (OBS), said correction (CORΦ) of said phase shift of said cosinusoidal and sinusoidal generated signals, and said first phase-locked loop (PLL1); b) it is determined whether said harmonics state observer (OBS) is converging; c) said cosinusoidal and sinusoidal generated signals are continuously corrected (CORH) by updating said determined harmonics when said harmonics state observer is converging; and d) said position and / or speed of said rotor is determined by means of the second phase-locked loop (PLL2) based on said corrected cosinusoidal and sinusoidal signals.

2. Method for determining the position and / or speed of a rotor according to Claim 1, wherein it is determined whether said harmonics state observer (OBS) is converging by checking whether the following equation is respected: y a − y ^ a 2 + y b − y ^ b 2 < ε with ya and yb said cosinusoidal and sinusoidal generated signals, respectively, ŷa and ŷb signals reconstructed by means of an estimated position and ε a predetermined threshold.

3. Method for determining the position and / or speed of a rotor according to either of Claims 1 and 2, wherein said first and second phase-locked loops (PLL1, PLL2) comprise a proportional-integral regulator (PI), and an integrator.

4. Method for determining the position and / or speed of a rotor according to Claim 3, wherein the transfer function of said first and second phase-locked loops (PLL1, PLL2) is written: θ ^ s θ s = 1 + K p K i ∗ s 1 + K p K i ∗ s + 1 K i ∗ s 2 with θ the position of said rotor, θ̂ an estimated position of said rotor, s the Laplace parameter, Kp the proportional coefficient of said proportional-integral regulator, and Ki the integral coefficient of said proportional-integral regulator.

5. Method for determining the position and / or speed of a rotor according to Claim 4, wherein said integral coefficient Ki of said first phase-locked loop (PLL1) is less than said integral coefficient Ki of said second phase-locked loop (PLL2).

6. Method for determining the position and / or speed of a rotor according to any of the preceding claims, wherein said cosinusoidal and sinusoidal signals are corrected (CORH) by filtering said determined harmonics, and possibly by correcting the phase shift of said cosinusoidal and sinusoidal signals.

7. Method for determining the position and / or speed of a rotor according to any of the preceding claims, wherein said harmonics state observer (OBS) implements a transfer function: y ^ k y = α ∗ s s 2 + k ∗ ω 2 1 + ∑ k = 0 → n α ∗ s s 2 + k ∗ ω 2 with s the Laplace parameter, α a gain, k an order of the harmonic in question, n a number of harmonics of said cosinusoidal and sinusoidal signals, ω the fundamental frequency, y the generated signal in question among the cosinusoidal and sinusoidal signals, and ŷk an estimated harmonic of order k of said generated signal in question among the cosinusoidal and sinusoidal signals.

8. Method for determining the position and / or speed of a rotor according to Claim 7, wherein said gain α is less than said fundamental frequency, and preferably said gain α is less than one tenth of said fundamental frequency ω.

9. Method for determining the position and / or speed of a rotor according to either of Claims 7 and 8, wherein said harmonics state observer (OBS) further determines fundamental coefficients of said harmonics and offsets of said cosinusoidal and sinusoidal generated signals.

10. Method for determining the position and / or speed of a rotor according to Claim 9, wherein said phase shift between said cosinusoidal and sinusoidal generated signals is identified by means of said fundamental coefficients of said harmonics using an arctangent function.

11. Method for determining the position and / or speed of a rotor according to Claim 10, wherein said phase shift Φ between said cosinusoidal and sinusoidal generated signals is determined by means of the equation: ϕ = atan L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ with L 1 b ∗ sin ϕ = − L ^ 11 b ∗ cos Δθ − L ^ 12 b ∗ sin Δθ L1b * cos(ϕ) = -L̂12b * cos(Δθ) - L̂11b * sin(Δθ), and with L1b a fundamental coefficient of one of said cosinusoidal and sinusoidal generated signals, L̂11b and L̂12b said fundamental coefficients determined by said harmonics state observer, Δθ a difference between the measured position of said rotor and the position of said rotor estimated by said first phase-locked loop (PLL1).

12. Method for determining the position and / or speed of a rotor according to any of the preceding claims, wherein said phase shift of one of said cosinusoidal and sinusoidal generated signals is corrected (CORΦ) by means of the equation y fb − correction = L 1 b ∗ y fb − y fa L 1 a ∗ L 1 b ∗ sin ϕ L 1 b ∗ cos ϕ with yfb-correction the fundamental of said corrected generated signal of one of said cosinusoidal and sinusoidal generated signals, L1a and L1b the fundamental coefficients of said cosinusoidal and sinusoidal generated signals, respectively, yfa and yfb the fundamentals of said cosinusoidal and sinusoidal measured generated signals, respectively, and Φ said phase shift.

13. Method for determining the position and / or speed of a rotor according to any of the preceding claims, wherein said position sensor is a magnetostrictive sensor, an inductive sensor, an encoder, a GMR sensor, an AMR sensor, a TMR sensor or a resolver.

14. Method for controlling an electric machine, said electric machine comprising a position sensor for sensing a position of the rotor of said electric machine, said position sensor generating a cosinusoidal signal and a sinusoidal signal, wherein the following steps are implemented: a) said position and / or speed of said rotor is determined by means of said method according to any of the preceding claims and of said signals generated by said position sensor; and b) said electric machine is controlled depending on said position and / or speed determined beforehand.