WO2005095896A1

WO2005095896A1 - Optical angular position sensor

Info

Publication number: WO2005095896A1
Application number: PCT/FR2005/000793
Authority: WO
Inventors: Mustapha Remouche; Ayoub Chakari; Patrick Meyrueis; Francis Georges
Original assignee: Universite Louis Pasteur, U.L.P.; BEI IDEACOD SAS (Société par Action Simplifiée)
Priority date: 2004-03-31
Filing date: 2005-03-31
Publication date: 2005-10-13
Also published as: WO2005095896B1; FR2868527B1; FR2868527A1

Abstract

The invention relates to an optical sensor (1) which can be used to detect, unambiguously and with high resolution, the absolute angular position over a full revolution, the number of revolutions performed and the direction of rotation of the rotating object independently of the speed of rotation. The inventive sensor comprises: a polarised light source (2) which illuminates a polariser (3) that rotates proportionally to the rotation of the object to be analysed, a detector (19) which captures the intensity of the beam passing through the polariser, and a signal analyser assembly (20). The invention also comprises a means for resolving indeterminacies (23) by generating a second measurement signal that is phase shifted in relation to the first signal, whereby said second phase-shifted signal can coexist with the first signal or replace same periodically. The invention is of interest to manufacturers and users of angular position sensors.

Description

ANGULAR POSITION OPTICAL SENSOR

The present invention relates to an optical sensor for measuring the angular position of an object rotating in a revolution. This sensor also makes it possible to determine the direction of rotation of the rotating object and the number of turns made by it. The invention also relates to an optical method for unambiguously determining the absolute angular position over a complete revolution, the number of revolutions carried out and the direction of rotation of a rotating object to be studied. In many situations, whether in industry or in everyday life, it is useful to know the speed and the angular position of a rotating object. For example, it is sometimes interesting or necessary to know the speed and the angular position of the shaft of an engine, a vehicle wheel, a rolling mill cylinder or any rotating mechanism of a Machine tool. Conventionally, when one wants to know the angular position of a rotating object, a position sensor of the incremental encoder or absolute encoder type is used exploiting optical, magnetic, mechanical or other phenomena. An incremental encoder, also called amplitude modulated optical relative encoder comprises a graduated disc at the periphery which rotates in connection with the object to be studied. The light emitted by a light-emitting diode is reflected by the graduations of the disc towards a receiving phototransistor which saturates and blocks at the rate of the scrolling of the graduations. The pulses are counted for give information concerning the angular displacement (number of pulses delivered from a home position) and / or information concerning the speed (number of pulses per unit of time). An incremental encoder generally has several channels: a channel giving one pulse per revolution used to count the number of revolutions performed, a channel giving n pulses per revolution and an additional channel identical to the previous one but whose signals are more or less phase shifted 90 ° according to the direction of rotation. It is thus also possible to obtain information concerning the direction of rotation of the object. However, such a sensor does not make it possible to know the exact position of the rotating object, but only its relative position relative to the initial position it occupied at the start of the measurement. When one wants to know the absolute position of the rotating object, one generally uses an optical sensor of the absolute encoder type with amplitude modulation. It is a disc divided into concentric tracks which rotates at the same time as the object to be studied. Each track comprises an alternation of reflecting or absorbing zones which divide the disc into angular sectors identifiable by the state of the zones of each track which are associated with it and which thus form a unique identification code. Thanks to this optical code, a row of emitting diodes placed in front of each track, associated with a row of corresponding receivers makes it possible at all times to identify the angular sector facing it and thus to give absolute position information for the rotating object. The number of tracks on the disc fixes the number of discrete positions that can be defined (n tracks corresponding to 2 ⁿ positions) and therefore determines the resolution of such an encoder. For obvious reasons related to its operating principle and to the space that can reasonably be accepted for such a sensor, its resolution is limited. With the most sophisticated sensors of this type, we cannot currently exceed a resolution of the order of 2 ¹⁶ . If a higher resolution is required, this type of sensor is unsuitable. To improve resolution and avoid the aforementioned drawbacks, a new type of angular position sensor has been developed in recent years. These are optical sensors operating on the principle of polarized light. Such a sensor has for example been described in patent application EP 0,246,892 in the name of BRITISH AEROSPACE PUBLIC LIMITED COMPANY. This optical sensor includes a fixed polarizer, a mobile polarizer in rotation actuated by a drive unit which rotates the mobile polarizer in proportion to the angular or linear displacement of the object to be studied, a light source, an optical fiber with polarization maintenance making it possible to direct a beam of polarized light of known intensity through the polarizers and an optical fiber with polarization maintenance placed at the output of these polarizers and making it possible to bring the outgoing light to a detector to measure its intensity. According to Malus' law, the rotation of the polarizer causes the intensity of the polarized light passing through it to be modulated. The rotation of the polarizer being proportional to the displacement of the object to be studied, this modulation translates the change in position of the object. Advantageously, this optical sensor no longer gives a result chosen from a set of discrete values, but belonging to a continuous range of values. However, the use of polarization-maintaining optical fibers which are very sensitive to vibrations is a source of errors which in another way limits the resolution that can be achieved. Furthermore, according to Malus' law, the measured intensity is a function in square cosine of the angular position θ of the rotating polarizer. This function is not bijective, for a measured intensity there are four possible corresponding angular position values. This prior sensor does not provide any means of lifting this indeterminacy, which is a very important limitation. In addition, this sensor does not make it possible to determine the direction of rotation of the rotating object. The object of the invention is to provide a high resolution optical sensor which makes it possible to unambiguously detect the absolute angular position over a complete revolution, the number of revolutions carried out and the direction of rotation of the rotating object independently of the speed of rotation. To solve this technical problem, the sensor according to the invention comprises a linearly polarized light source emitting a first measurement beam which will pass through a mobile polarizer in rotation driven in proportion to the movement of the object to be studied, a detector detecting the beam intensity at the output of the polarizer and transmitting it as an electrical signal to a signal analyzer assembly. Advantageously, the light beam is divided at the entrance to the assembly so as to form a reference beam not passing through the polarizer, the intensity of which is measured and transmitted to the signal analyzer assembly so as to eliminate any fluctuation linked to the light source. According to an essential characteristic of the invention, the sensor further comprises a means for lifting the indeterminacies, preferably comprising a delay blade, which generates from the first measurement beam, a second measurement beam out of phase with the first measurement beam . According to the variants envisaged, this second phase-shifted measurement beam can coexist permanently with the first measurement signal or replace the first measurement beam periodically and alternately. Signal processing can be carried out in several ways, in particular according to a purely optoelectronic variant (sensor with a single axis) or according to an opto-mechanotronic variant (sensor with two axes). Advantageously, the resolution and the accuracy of the sensor according to the invention are independent of the speed of rotation, which is not the case for other amplitude modulated optical sensors. The invention also provides an optical method for unambiguously determining the absolute angular position over a complete revolution, the number of revolutions carried out and the direction of rotation of a rotating object to be studied. This process includes the following steps: - generation of a linearly polarized light beam; - creation from this polarized light beam of a first measurement beam and a second measurement beam out of phase with the first, these two measurement beams existing simultaneously or alternately;

- crossed by the first and second measurement beams, a mobile polarizer in rotation, driven in proportion to the movement of the object to be studied;

- detection of the intensity of the first and / or of the second measurement beam at the output of the polarizer;

- calculation of the desired results among the absolute angular position, the direction of rotation and the number of revolutions made by the object to be studied, from the value (s) of intensity detected. Preferably, the method according to the invention also includes the following steps:

- Creation from the polarized light beam of a reference beam not crossing the mobile polarizer; - detection of the intensity of the reference beam. According to a preferred variant of the invention, the optical method can also include the following steps:

- modulation of the light intensity of the linearly polarized light beam;

- determination of the absolute angular position of the object to be studied from:. of the calculation of the value of the real transfer function of the polarizer corresponding to the slope of the line defined by at least two pairs of detected intensity values, couples formed by the intensity of a measurement beam and of the intensity of the reference beam, and. a transfer table grouping together the slope / corresponding angular position pairs. Other features and benefits of the invention will appear on reading the detailed description which follows, description made with reference to the accompanying drawings, in which: Figure 1 is a very general block diagram of the angular position sensor according to the invention; . Figure 2 is a simplified perspective diagram of a first variant with a single axis of the part of the angular position sensor according to the invention comprising the rotating polarizer; . Figure 3 is a simplified perspective diagram of a second two-axis variant of the part of the angular position sensor according to the invention comprising the rotating polarizer; . Figure 4 is a simplified plan diagram of a third two-axis variant of the angular position sensor according to the invention; . Figure 5 is the intensity curve obtained from the assembly of Figure 4; . FIG. 6 is a block diagram of an angular position sensor according to the invention comprising means for lifting the indeterminacies by generation of a second phase-shifted measurement signal permanently coexisting with the first measurement signal; . Figure 7 is a simplified diagram of a first embodiment of a single-axis angular position sensor according to the principle of Figure 6; . Figure 8 is a simplified diagram of a second embodiment of a single-axis angular position sensor according to the principle of Figure 6; . Figure 9 is the intensity curve obtained from the assemblies of Figures 7 and 8; . Figure 10 is a curve representing the signal two-state square resulting from a U-turn detector; . FIG. 11 is a block diagram of the signal processing analyzer assembly according to the principle of the assemblies of FIGS. 7 and 8; . Figure 12 is a simplified diagram of an exemplary embodiment of an angular position sensor with two axes according to the principle of Figure 6; . Figure 13 is the intensity curve obtained from the assembly of Figure 12 for a gear ratio of 1/4; . Figure 14 is the intensity curve obtained from the assembly of Figure 12 for a gear ratio of 1/2; . FIG. 15 is a block diagram of an angular position sensor according to the invention comprising means for lifting the indeterminacies by generation of a second phase-shifted measurement signal periodically replacing the first measurement signal; . FIG. 16 is a simplified diagram of an exemplary embodiment of an angular position sensor with a single axis according to the principle of FIG. 15; . Figure 17 is the intensity curve obtained from the assembly of Figure 16; . Figure 18 is a simplified diagram of an exemplary embodiment of an angular position sensor with two axes according to the principle of Figure 15; . FIG. 19 is the intensity curve obtained from the assembly of FIG. 18. The optical angular position sensor according to the present invention will now be described in detail with reference to FIGS. 1 to 19. The equivalent elements represented on the different figures will bear the same reference numbers. The optical sensor 1 according to the invention comprises a light source 2 which generates a linearly polarized light beam of intensity I ₀ . Different types of light sources can be used, such as a linearly polarized laser source, a polarized diode, a non-polarized diode followed by a fixed polarizer or the like. A laser source is preferably used which generates a coherent beam of light and linearly polarized, for example of wavelength λ = 633 nm. This light source 2 comes to light a polarizer 3 mobile in rotation and whose movement is linked to that of a rotating object 4 whose rotation we want to study, that is to say whose angular position we want to know at all times θ _ob j _e t _f the direction of rotation and the number of turns it has already made. This rotating object can be arbitrary. As an example of application, there may be mentioned the shaft of an engine, a vehicle wheel, a rolling mill cylinder, a rotating mechanism of a machine tool, or any other rotating object of which one wishes to know precisely the angular position. It is also conceivable that the angular position sensor 1 according to the invention is used to measure the linear displacement of any object. The linear movement of the object to be studied is then transformed into a rotational movement by any appropriate means easily achievable by those skilled in the art. In this case, the rotating object 4 shown in the different figures corresponds to the rotating part of the movement transformation means. In order to ensure correct operation of the angular displacement sensor 1 according to the invention, the optical part thereof is isolated from light room. It is thus for example enclosed in an opaque housing 5 of any suitable shape, which is shown schematically and by way of example in certain figures. As can be seen in Figures 2 and

3, the polarizer 3 is preferably in the form of a thin polymer plate and preferably in the form of a disc 6. The polarizer 3 is mounted on an input rod 7 used to transmit the rotational movement of the rotating object 4 at the polarizer 3. The input rod 7 can be connected to the rotating object 4 directly as shown diagrammatically in FIG. 2 or by means of elements which do not modify the rotational movement. The angular displacement θ of the rotating object 4 is thus transmitted without modification to the polarizer 3 which rotates in the same way and at the same speed as the object 4. Thus for example, in the context of a variant only optoelectronic to a single axis shown in Figure 2, the polarizer 3 is mounted on an input rod 7 which can be fixed on the housing 5 using one or more ball bearings 8 or any other equivalent means ensuring good measurement stability. The rotating object 4 can be directly linked to the rod 7 or be secured to a rod 9 coupled to the input rod 7 by a coupling system 10. The coupling system 10 and the ball bearing 8 allow 'Move away, if necessary, the rotating object 4 from the optical part of the sensor 1 and ensure better mechanical stability of the sensor during operation. The attachment of the polarizer 3 to the input rod 7 can be done by any suitable means and for example by means of a central fixing element. or as shown using two support discs 11 of diameter smaller than the polarizing disc 6, fixed to the end of the rod 7 on either side of the polarizer and leaving free on the polarizer a peripheral ring 12 intended to be crossed by the beam of polarized light. A transmission system 13 proportionally modifying the rotational movement can also be interposed between the rotating object 4 and the input rod 7. This is for example a transmission system 13 by gears or by belts. By way of example, FIGS. 3 and 4 show two variants comprising such a transmission system 13 by gear. These variants are called opto-mechanotronic (angular position sensor with two axes). A transmission rod 14, which can be fixed to the housing 5 using for example one or more ball bearings 8, carries a toothed wheel 15. As indicated previously, this transmission rod can be secured to the object turning 4 directly (Figure 3) or for example via a coupling 10 and an additional rod 9 (Figure 4). The toothed wheel 15 is engaged with a toothed wheel 16 which, according to the variant envisaged in FIG. 3, is carried by the input rod 7, movable in rotation, on which the polarizer 3 is mounted. In this case, the beam of polarized light emitted by the light source 2 must pass through the free peripheral ring 12 of the rotating polarizer. In the embodiment shown in Figure 4, the toothed wheel 16 is integral with the movable outer ring of a ball bearing 17 carried by the input rod 7 which in this case is fixed and hollow. Advantageously, the polarizer 3 can be directly fixed to the toothed wheel 16 or the mobile crown of the ball bearing 17. The polarized light source 2 is positioned in the axis of the hollow input rod 7, or even at the inside of it. In this case, the polarized light beam passes through the free central part 18 of the mobile polarizer, which also makes it possible to minimize the optical noise. The angular displacement θ of the rotating object 4 is transmitted to the polarizer 3 according to a transmission ratio depending on the gear ratio between the two toothed wheels 15 and 16. This transmission ratio can be arbitrary, but it is advantageously chosen with a value of 1/4, as shown, for a reason which will be explained later. In this particularly advantageous case, when the rotating object 4 rotates, the polarizer 3 rotates a quarter of a turn. Such a configuration allowing the light beam to pass through the central part 18 of the rotating polarizer 3 is obviously not limited to a two-axis assembly, but could also be adapted to a single-axis assembly. When the angular position sensor 1 according to the invention is in operation, the light source 2 emits a polarized light beam of intensity I ₀ which passes through the polarizer 3. The rotation of the polarizer 3, caused by and proportional to the rotation of the rotating object 4, induces a variation in the light intensity at the output of the polarizer 3. According to Malus' law, the intensity I of the beam at the output of the rotating polarizer depends on the angular displacement θ produced by the polarizer 3 and corresponds to the following function: I = I _m cos ² θ. In the case of a one-axis assembly such as that of Figure 2, θ = θ _object and in the case of a two-axis assembly with transmission system such as those of Figures 3 and 4, θ is proportional at θ _{ob: jet} . The outgoing light beam constitutes the measurement signal which is picked up by a measurement signal detector 19 placed opposite. The measurement signal detector 19, preferably a photodiode, detects the intensity I of the beam and transmits the detected value in the form of an electrical signal to an analyzer assembly 20. The angular motion sensor 1 according to the invention is designed to be able to be used in industrial units, vehicles, or other applications for which it is subjected to conditions which are not optimal laboratory conditions. Thus, for example, it must sometimes withstand vibrations, high temperatures or other disturbances which must not affect the reliability and the resolution of the measurements carried out. To eliminate the influence on the results of any fluctuation in the light source which could cause annoying variations in the initial light intensity, it is advantageous to make a reference loop. For this, the polarized light beam is divided by a beam splitter 21, placed for example at the input of the assembly, so as to form a measurement beam and a reference beam not passing through the polarizer. The beam splitter 21 does not modify the polarization state of the beams and can for example be made up of a separating cube, a semi-transparent mirror or the like. The intensity I _R of the reference beam is picked up by a reference signal detector 22, preferably a photodiode, and transmitted in electrical form to the signal analyzer assembly 20. Thanks to this reference loop, the instantaneous value of l initial intensity of the light source 2 is permanently known and transmitted to the analyzer assembly 20 which can thus overcome any variation thereof, for example by dividing the intensity detected by the detector (s) measurement signal 19 by the reference intensity I _R detected by the reference detector 22. The signal analyzer assembly 20 is an electronic processing module which comprises several stages and in particular, according to a preferred variant, a stage of amplification used to amplify the signals from the various detectors 19 and 22, a stage for dividing the measurement signals by the reference signal and a stage for calculating the desired results among the angular position, the direction of rotation and the number of turns. The analyzer assembly 20 may further comprise, depending on the case, a device for counting turns, half-turns or quarter-turns. The calculation stage of the signal analyzer assembly 20 calculates an angular displacement value θ from the detected value of cos ² θ. However, and as already indicated in the introduction, this function, not being bijective, generates indeterminations as to the real value of the angular displacement θ. The embodiments illustrated in FIGS. 3 and 4 make it possible to resolve the indeterminacies in a revolution when the transmission ratio is chosen equal to 1/4. The curve intensity obtained at the output of the rotating polarizer

3 with this type of assembly is given in FIG. 5. When the object to be studied 4 travels a full revolution, the polarizer makes only a quarter turn. The cos ² fonction function being bijective on a quarter turn of the polarizer, the calculation stage of the analyzer assembly

20 easily determines the angular position of the polarizer θ in the first quarter-turn which is equal to θ = arccos (- Vha). Consequently the real angular displacement of the rotating object 4 in the first turn is equal to four times θ, that is to say, θ ₀ bjet = 4 arccos (-yjl / lm). In the second round, the intensity through the rotating polarizer given by the Malus law becomes: I = I _m (l-cos ² θ) = I _m sin θ. In this case, the angular position of the object 4 is determined by the relation: θ _object = 4 arcsin (Vl / Im). However, when the rotating object 4 can change the direction of rotation, there is still an indeterminacy to determine the direction of rotation of the rotating object at the minima and maxima of the intensity curve measured in cos ² θ . In order to solve these problems and according to an essential characteristic of the invention, the angular displacement sensor 1 according to the invention further comprises a means for lifting the indeterminacies 23. This means 23 operates by generation of a second phase-shifted measurement signal with respect to the first measurement signal. According to a first embodiment shown in the block diagram of FIG. 6, this first measurement signal and this second measurement signal out of phase with the first can coexist permanently. To do this, the incident light beam is divided into two measurement beams using a beam splitter 21 which has no influence on the state of polarization of the beams and which can for example be a separating cube, a semi-transparent mirror or the like. In the preferential case where the sensor according to the invention comprises a reference loop, the beam splitter 21 used to create the reference beam can be placed before (FIG. 6) or after (FIGS. 7 and 8) that used to create the second measurement beam. These two beam splitters 21 can be identical or different. The first measurement beam passes through the rotating polarizer which modifies its intensity according to the Malus law I _x = I _ml cos ² θ and reaches a first measurement signal detector 24. This first measurement signal detector 24 detects the intensity I _x and transmits the detected value in the form of an electrical signal to the analyzer assembly 20. The second measurement beam encounters a means for generating a phase shift 25 which creates an angle phase shift φ with respect to the first measurement beam. The second measurement beam, phase shifted with respect to the first, also crosses the rotating polarizer which modifies its intensity according to the law of Malus I ₂ = I _m2 cos ² (θ-φ). It then leads to a second measurement signal detector 26, which detects the intensity I ₂ and transmits it to the analyzer assembly 20. The two measurement signal detectors 24 and 26 may be identical or different and are preferably photodiodes . The means for generating a phase shift 25 can be arbitrary as long as it does not modify the state of polarization of the beam. The phase shift can be chosen from any angle φ between 0 and 90 °. Advantageously, it can be chosen equal to 45 °. Two variants have for example been imagined in FIGS. 7 and 8 for mounting with a single axis. A first variant represented in FIG. 7 consists in geometrically creating the phase shift by shifting by an angle φ the point of impact 27 of the second intensity measurement beam I ₂ relative to the point of impact 28 of the first measurement beam of intensity I ₁ on the rotating polarizer 3. A second variant illustrated in FIG. 8 uses a retarder plate 29. In this case, the production of the phase shift φ between the two measurement beams comprises two stages. The first consists in obtaining a second measurement beam in phase with the first. For this, the point of impact 27 of the second beam is chosen so as to be arranged geometrically on the same diametral line as the point of impact 28 of the first beam. Two examples of possible arrangements are given in FIG. 8. The two points of impact can thus for example be arranged on the peripheral ring 12 of the polarizer 3 in a diametrically opposite manner (180 ° offset) or not. The second step consists in interposing a delay plate 29 on the path of the second measurement beam before it crosses the rotating polarizer 3. This delay plate 29 is preferably a half-wave plate (λ / 2) oriented at an angle α with respect to the direction of polarization. This plate does not modify the polarization state of the beam but creates a phase shift φ = 2α at the output of the polarizer 3 between the two measurement beams. Advantageously, an orientation of the blade half-wave (λ / 2) with an angle α of 22.5 ° relative to the direction of polarization makes it possible to obtain a chosen phase shift of φ = 45 ° between the two measurement beams. Preferably, it is possible to use a retarding strip 29 which is in the form of a film. The intensity values l _x and I ₂ transmitted to the analyzer assembly 20 by the measurement signal detectors 24 and 26 can be represented as a function of the angle of rotation of the rotating object in the form of two offset curves. by an angle φ corresponding to the phase shift. An example of such a pair of 45 ° phase-shifted curves has been shown in FIG. 9. The analyzer assembly 20 receives at each instant the two detected values of l _x and I ₂ . It therefore has a system of two equations with a single unknown that it can easily solve using simple and common electronic components. The second equation allows him to remove any uncertainty as to the real value of θ and the direction of rotation. Advantageously, the analyzer assembly 20 in this case comprises a half-turn detector 30 allowing it to know in which half-turn of its angular displacement is the polarizer 3. An example of very simple embodiment of such a detector 30 has has been shown in Figures 7 and 8. However, any other half-turn detection means can be used satisfactorily, in particular a magnetic detector or the like. In the example shown, an absorbent layer 31, preferably of width substantially equal to 5 mm, has been etched on the periphery of a first half-turn of the polarizer 3. A light source for example of the diode type electroluminescent, not shown in the figures, is arranged facing the polarizer at a level corresponding to the absorbing zone 31. This source emits a light beam which, when the absorbing layer 30 is absent, crosses the rotating polarizer 3 and is picked up by a detector signal arranged opposite, for example of photodiode type and also not shown. When this layer is located between the light source and the detector, it prevents the beam from passing through the polarizer 3. Consequently, the U-turn detector 30 generates a square signal with two logic states 0 or 1 as shown on the curve of Figure 10. If the U-turn detector indicates a logic zero state, the polarizer 3 is in the first U-turn of its angular displacement. In this case, the angular position θ of the polarizer 3 is determined by the relation: θ = (l / 2) arctan (I _ml I ₂ / I _m2 I ₁ ). In the case where the detector indicates a logic state 1, the polarizer 3 is in the second half-turn of its angular displacement and its real angular position is calculated by the following relation: θ = 180 ° + (l / 2) arctan (I _ral I ₂ / I _m2 I ₁ ). A block diagram of a preferred embodiment of the signal analyzer assembly 20 corresponding to this variant of the sensor 1 according to the invention has been given in FIG. 11. The signal analyzer assembly 20 represented in FIG. 11 comprises a high frequency filtration stage 32 used to eliminate the continuous component of the measurement signals I _x and I ₂ coming from the detectors 24 and 26, an amplification stage 33 serving to amplify the signals I _x and I ₂ coming from the filtration stage 32 and the reference signal from detector 22, a signal division stage 34 and a calculation stage desired results 35 among the angular position, the direction of rotation and the number of turns performed. The elimination of the DC component of the measurement signals I _x and I ₂ makes it possible to directly determine the angular position θ and the direction of rotation in the first half-turn of the polarizer 3 between 0 ° and 180 °, because the ratio I _ml I ₂ / I _m2 I ₁ is equal to the tangent of 2θ (tan2θ = I _ml I ₂ / I _m2 I ₁ ) and therefore θ = (l / 2) arctan (I _ml I ₂ / l _m2 I ₁ ). In the practical case, it is preferable that the two amplitudes I _ml and I _m2 are equal (I _ml = I _m2 ). In this case, the angular position of the rotating object 4 is given by: θ _object = (l / 2) atan (I ₂ / I ₁ ). Obviously, the means for lifting the indeterminacies 23 by generation of a second measurement signal out of phase with the first measurement signal can also be adapted to the two-axis variant of the sensor according to the invention. In accordance with the embodiment represented in the block diagram of FIG. 6, the two measurement signals coexist permanently and the second measurement beam encounters a means of generating a phase shift 25 which creates a phase shift of angle φ relative to at the first measurement beam. The two variants of the phase-shift generation means 25 described above (geometric phase shift or using a retarding blade 29) are also adaptable to mounting with two axes. An embodiment with a retarder blade

29 for an assembly with two axes has been shown by way of example in FIG. 12. As previously, the creation of the phase shift is carried out in two stages: obtaining two beams in phase by geometrically positioning the points of impact 27 and 28 of the two beams on the same diametrical straight line of the rotating polarizer 3 (two positioning examples are illustrated in FIG. 12) and interposition of a retarder plate 29 on the path of the second measurement beam before it crosses the polarizer 3. The half-wave retarder plate (λ / 2) can advantageously be oriented with an angle α of 22.5 ° relative to the direction of polarization, so as to generate a chosen phase shift of 45 °. The intensity curves I _λ and I ₂ obtained at the output of the polarizer 3 can be represented as a function of the angle of rotation of the rotating object in the form of two curves offset by an angle φ corresponding to the phase shift. An example of such a pair of 45 ° phase shifted curves has been shown in FIG. 13. The analyzer assembly 20 receives at each instant the two detected values of I _x and I ₂ . It therefore has a system of two equations with a single unknown that it can easily solve using simple and common electronic components. The first equation, I _x = I _ml cos ² θ, makes it possible to directly determine the real angular position of the rotating object 4 using the relation: θ _object ⁼ 4 arccos (Vl / I) or the relation θ _object ⁼ 4 arcsin (VlIm). The first equation also makes it possible to determine the direction of rotation between the characteristic points (the minima or the maxima of To remove any uncertainty concerning the direction of rotation of the sensor, each time the intensity value of the first measurement signal I passes by a characteristic point, it suffices to check the direction of the variation of the intensity of the measurement signals I _x and I _2. If the direction of variation is the same for the two signals (increasing or decreasing), the sensor rotates clockwise. In the opposite case, that is to say the signal I _x increasing and I ₂ decreasing or reverse, the sensor rotates in the anticlockwise direction. If the gear ratio is 1/2, the rotating object rotates one full revolution when the polarizer rotates half a revolution. In this case, the permanent presence of a second measurement signal I ₂ out of phase with the first measurement signal I _lf makes it possible to remove the indeterminacies and to directly determine the angular position and the direction of rotation of the rotating object 4 If the two signals are normalized and 45 ° out of phase, eliminating their continuous components allows their expressions to be rewritten as follows: I _x = cos (2θ) = cos (θ _object ) and I ₂ = sin (2θ ) = sin (θ _object ) where θ is the angular position of the rotating polarizer 3 and θ _object is the angular position of the rotating object 4 (2Θ = θ _object ). The angular position of the rotating object is calculated thus directly using the tangent function tan (θ _item) = L ^~ _2/1. ₁ and therefore θ _object = arctan (I ₂ / I ₁ ). The variation curves of I _x and I ₂ , for a revolution of the rotating object, are represented in FIG. 14. According to another embodiment represented in the block diagram of FIG. 15, the second measurement signal out of phase with the first, can replace the first measurement signal periodically and alternately. In this case, a periodic means for generating a phase shift 36 is placed on the path of the measurement beam before it crosses the rotating polarizer 3. This means 36 does not modify the state of polarization of the beam but adds periodically a phase shift of angle φ to the beam of measured. Unlike the previous embodiment, the two phase shifted measurement beams do not exist simultaneously but are replaced alternately and periodically. The measurement beam then crosses the rotating polarizer 3 which modulates its intensity. As before, the beam is then picked up by a measurement signal detector 12 and the value of its intensity is transmitted to the signal analyzer assembly 20. This is equal to I _x = I _m cos ² θ when the means periodic phase shift generation 36 does not create a phase shift and at I ₂ = I _m cos ² (θ-φ) when the periodic phase shift generation means 36 creates a phase shift angle φ. If we draw a curve representing the variation of the intensity at the output of the polarizer 3, that is to say corresponding that measured by the detector 19, as a function of the angle of rotation of the rotating object, a composite curve is obtained, similar to those shown in FIGS. 17 and 19, formed of portions of curve which follow one another periodically and corresponding respectively to I _x or to I ₂ . The phase shift can be chosen from any angle φ preferably between 0 and 90 °. Preferably, it is possible to use a phase shift of 90 ° as illustrated in the curves of FIGS. 17 and 19. This embodiment of the optical sensor 1 according to the invention can be indifferently adapted to mounting on one axis or to mounting in two axes. An exemplary embodiment of each of these arrangements has been shown respectively in FIGS. 16 and 18. The period characterizing the alternation of the beams is dependent on the rotational movement of the rotating polarizer 3. It is preferably chosen equal to a quarter of a turn of the polarizer 3, which corresponds to a quarter of a turn of the rotating object 4 for the assembly of FIG. 15 and to a complete revolution of the rotating object for the assembly of FIG. 18 which a preferred gear ratio equal to 1/4. The periodic change of state of the phase shift generation means 36 can be controlled by the analyzer assembly 20 by means of a feedback loop dependent on the value of the intensity sensed by the detector 19, so as to cause this change of state each time the intensity value passes through a characteristic point, for example and preferably each time it reaches a maximum or a minimum. Of course and as indicated above, the measurement beam can, before meeting the periodic means for generating a phase shift 36, be divided to form a reference beam. According to the variants shown in FIGS. 16 and 18, the periodic means for generating a phase shift 36 can be a movable retarding blade 37 which periodically comes into the path of the measurement beam so as to generate when it is present an angle phase shift φ. Preferably, it may be a half-wave retarding plate (λ / 2) oriented with an angle α for example equal to 45 ° relative to the direction of polarization, thus generating a chosen phase shift of 90 ° when it is crossed by the measurement beam. The movement of this movable blade 24 is preferably controlled by an OFF / ON shutter controlled by the analyzer assembly 13. Again, it is preferable to use a retarder blade 37 in the form of a film. In such measurement conditions, the composite curve of FIG. 17 is obtained for the mounting with an axis of FIG. 16 and the composite curve of FIG. 19 for the mounting with two axes of FIG. 18. Advantageously, these curves vary always in the same direction as long as the direction of rotation of the polarizer 3 remains unchanged and are reversed if this direction of rotation changes. It is therefore extremely easy to know the direction of rotation of the rotating object 4. In addition, it is sufficient to couple, within the analyzer assembly 20, the calculation stage with a quarter-turn counter device in order to be able to unambiguously give the angular position of the rotating object 4. The optical sensor 1 according to the invention is particularly advantageous because it makes it possible to give in a simple, reliable and precise manner both the absolute angular position, the direction of rotation and the number of turns performed by a rotating object. It is not limited in number of turns and works both at high and at low speed. It remains reliable and precise even when used in unfavorable conditions such as industrial operating conditions. In addition, the optical detector according to the invention is particularly tolerant with respect to off-center or misalignment. Obviously, the invention is not limited to the preferred embodiments described above and shown in the different figures, the person skilled in the art can make numerous modifications and imagine other variants without departing from the scope of the invention. One can thus cite by way of example a method of exploiting the measurement signals which makes it possible to obtain experimentally and without complicated calculation the value of the angular position θ of the rotating polarizer 3, linked to that θ _0bj e _t of the rotating object to be studied, starting from the real transfer function of the rotating polarizer 3. The real transfer function of the polarizer represents the real experimental response of the sensor which approaches the theoretical response in cos ² θ explained above. This function performs the correspondence between the measured signal and the value of the angular position. It is calculated using a basic principle consisting in modulating the light intensity of the measurement signal and that of the reference signal by varying the control signal of the light source used, for a known fixed angular position θ of the rotating polarizer 3 and to study the variation of the measurement signal as a function of the reference signal (translating that of the light source). The variation of the measurement signal I _m as a function of the reference signal I _R is a straight line characteristic of a given angular position θ, the equation of which is as follows: I _m (I _R ) = A * I _R + B in which A and B are constants which depend on the angular position θ. This equation can be rewritten as follows: Im (θ) = A (θ) * I _R + B (θ) A (θ) represents the slope of this line and depends only on the ratio of the gains of the amplifiers which is a constant of the system and the angular position θ. B (θ) represents the intersection with the measurement axis. It depends on the angular position θ, on the derived from electronics (offset) and parasitic reflections on the various diopters of the systems. For each angular position θ of the rotating polarizer 3, a different straight line is obtained, the value of the slope A of which can be determined experimentally. The variation of this slope A as a function of the angular position θ of the rotating polarizer constitutes the real transfer function of the polarizer. The fact that this transfer function depends only on A (θ) makes the measurement insensitive to offset electronics drift and takes into account the specific characteristics of the polarizer (transmission coefficient, reflection coefficient, etc.). This operating method requires a prior calibration step of the sensor, during which a transfer table is established which is stored in a memory, for example a ROM memory, of the analyzer assembly 20 of the sensor. This transfer table groups together thousands of pairs of slope A / corresponding angular position θ and is established experimentally for a given sensor. It is carefully established during a preliminary calibration phase which can advantageously be automated, during which several thousand measurements are carried out under optimal signal / noise conditions. Subsequently in use, to determine an angular position θ of the polarizer, the value of the intensity of the measurement signal I _m is detected by the measurement signal detector 19 and that of the reference signal I _R by the signal detector reference 22 for at least two different intensity values of the light source 2. The analyzer assembly 20 then calculates the slope A of the line defined by these at least two pairs of values (I _ra , I _R ), thus determining the value of the real transfer function of the polarizer for the angular position θ to be identified. The transfer table, stored in the analyzer assembly 20, then gives the angular position θ corresponding to this value of the transfer function of the polarizer. This method of processing the measurement signals can be used in all of the previously described variants of the sensor according to the invention. It can be applied indifferently to one or the other of the measurement signals, or to the two signals simultaneously or successively. This process is particularly advantageous because it is not sensitive to long-term drifts in electronics. In addition, it takes into account the manufacturing defects of the polarizer and of the associated optical elements. It thus makes it possible to use an inexpensive polarizer and standard elements. The use of the transfer table avoids any complicated calculation and achieves the desired resolution, precision being limited only by the associated measurement electronics. This operating process is fast, precise and allows a significant manufacturing dispersion of the optical system. In addition, it is perfectly suited to industrial operating conditions because it is not very sensitive to vibrations. This process can advantageously be used when the polarizer rotates slowly or when it stops and you want to know its angular position with great precision. It gives a more precise result than the use of theoretical mathematical formulas because it takes into account actual sensor faults. When the polarizer rotates at high speed, the mathematical methods previously described will then preferably be used, although the use of this operating method remains conceivable. It should be clearly understood that this advantageous method is only an example of a method for exploiting the signals detected by the sensor according to the invention, numerous other methods which can be perfected by those skilled in the art in the context of the invention.

Claims

1. Optical sensor making it possible to unambiguously detect the absolute angular position over a complete revolution, the number of revolutions made and the direction of rotation of a rotating object (4) to be studied, characterized in that it comprises:

- a laser source (2) generating a beam of coherent light and linearly polarized, which forms a first measurement beam;

- a means for lifting the indeterminacies (23) which generates, from the first measurement beam, a second measurement beam out of phase with the first measurement beam; - a mobile polarizer in rotation (3), driven in proportion to the displacement of the object to be studied (4) and crossed by the first and the second measurement beam;

- at least one measurement signal detector (19, 24, 26) measuring the intensity of a measurement beam at the output of the polarizer (3);

- a signal analyzer assembly (20) which calculates the desired results from the absolute angular position, the direction of rotation and the number of revolutions made by the object to be studied (4), from the value (s) intensity detected and transmitted by the measurement signal detector (s) (19, 24, 26).

2. Optical sensor according to the preceding claim, characterized in that the means for lifting the indeterminacies (23) comprises a beam splitter (21) and a means for generating a phase shift. (25) making it possible to create a second measurement beam, phase shifted by an angle φ relative to the first measurement beam and permanently coexisting with this first beam.

3. Optical sensor according to the preceding claim, characterized in that the second measurement beam is 45 ° out of phase with respect to the first measurement beam.

4. Optical sensor according to claim 2 or 3 characterized in that the means for generating a phase shift (25) is a geometric means consisting in geometrically shifting by an angle φ the point of impact (27) of the second beam of measurement with respect to the point of impact (28) of the first measurement beam on the mobile polarizer (3).

5. Optical sensor according to claim 2 or 3 characterized in that the means for generating a phase shift (25) comprises a delay blade (29).

6. Optical sensor for unambiguously detecting the absolute angular position over a complete revolution, the number of revolutions made and the direction of rotation of a rotating object (4) to be studied, characterized in that it comprises: - a source of linearly polarized light (2) which emits a first measurement beam;

- a means for lifting the indeterminacies (23) comprising a retarding plate (29, 37) which generates from the first measurement beam a second measurement beam out of phase with the first measurement beam;

- a mobile polarizer in rotation (3), driven in proportion to the displacement of the object to be studied (4) and crossed by the first and the second measurement beam;

- a signal analyzer assembly (20) which calculates the desired results from the absolute angular position, the direction of rotation and the number of turns carried out by the object to be studied (4), from the intensity value (s) detected and transmitted by the measurement signal detector (s) (19, 24, 26). . Optical sensor according to the preceding claim, characterized in that the retarding blade

(29, 37) is in the form of a film. 8. Optical sensor making it possible to unambiguously detect the absolute angular position over a complete revolution, the number of revolutions made and the direction of rotation of a rotating object (4) to be studied, characterized in that it comprises:

- a linearly polarized light source (2) which emits a first measurement beam; a means for lifting the indeterminacies (23) comprising a periodic means for generating a phase shift (36) which generates from the first measurement beam a second measurement beam out of phase with the first measurement beam, replacing periodically and alternately this first beam;

- a signal analyzer assembly (20) which calculates the desired results from the absolute angular position, the direction of rotation and the number of revolutions made by the object to be studied (4), from the value (s) intensity detected and transmitted by the measurement signal detector (s) (19, 24, 26). . Optical sensor according to claim previous characterized in that it comprises a periodic means for generating a phase shift (36) which generates from the first measurement beam a second measurement beam out of phase with the first measurement beam, periodically and alternately replacing this first beam according to a period depending on the rotational movement of the mobile polarizer (3). 10. Optical sensor according to the preceding claim characterized in that the alternation period of the beams is equal to a quarter turn of the mobile polarizer (3). 11. Optical sensor according to any one of claims 8 to 10 characterized in that the second measurement beam is phase shifted by 90 ° relative to the first measurement beam. 12. Optical sensor according to any one of claims 8 to 11 characterized in that the periodic means for generating a phase shift (36) is a movable retarding blade (37) which periodically comes into the path of the beam of measurement and whose movement is controlled by an OFF / ON shutter controlled by the analyzer assembly (20). 13. An optical sensor according to any one of claims 6 to 12 characterized in that the polarized light source (2) is a laser source generating a coherent light beam and linearly polarized. 14. Optical sensor according to any one of the preceding claims, characterized in that it further comprises a reference loop in which the first measurement beam is divided by a beam splitter (21) so as to form a reference beam not crossing the mobile polarizer (3), the intensity of which is picked up by a reference signal detector (22) and transmitted to the signal analyzer assembly (20). 15. Optical sensor according to claim 2 or 14 characterized in that at least one of the beam splitters (21) is a separating cube or a semi-transparent mirror which has no influence on the polarization state of the beams . 16. Optical sensor according to any one of the preceding claims, characterized in that the rotating polarizer (3) is in the form of a thin disc (6) made of polymer. 17. Optical sensor according to any one of the preceding claims, characterized in that the rotating mobile polarizer (3) is linked to the rotating object (4) directly or by means of elements transmitting the rotational movement without modification. from the rotating object (4) to the polarizer (3). 18. Optical sensor according to any one of claims 1 to 16 characterized in that the rotating mobile polarizer (3) is linked to the rotating object (4) via a transmission system (13) modifying proportionally the rotational movement of the rotating object (4) to be studied. 19. Optical sensor according to the preceding claim characterized in that the transmission ratio of the transmission system (13) is equal to 1/4 or 1/2. 20. Optical sensor according to any one of the preceding claims, characterized in that at least one of the signal detectors (19, 22, 26, 28) is a photodiode. 21. Optical sensor according to any one of the preceding claims, characterized in that the signal analyzer assembly (20) comprises a device for counting turns, half-turns or quarter turns. 22. Optical sensor according to claim 14 characterized in that the signal analyzer assembly (20) is an electronic processing module with several stages comprising in particular a stage for amplifying the signals coming from the various detectors (19, 22, 26, 28), a stage for dividing the measurement signals by the reference signal and a stage for calculating the results. 23. Optical sensor according to any one of the preceding claims, characterized in that it further comprises an opaque housing (5) isolating the optical part of the sensor from ambient light. 24. Optical method for unambiguously determining the absolute angular position over a complete revolution, the number of revolutions carried out and the direction of rotation of a rotating object (4) to be studied, characterized in that it comprises the following steps:

- generation of a linearly polarized light beam;

- Creation from this polarized light beam of a first measurement beam and a second measurement beam out of phase with the first, these two measurement beams existing simultaneously or alternately;

- crossed by the first and second measurement beams, a rotating polarizer (3), driven in proportion to the movement of the object to be studied (4); - detection of the intensity of the first and / or of the second measurement beam at the output of the polarizer (3);

- calculation of the desired results among the absolute angular position, the direction of rotation and the number of turns made by the object to be studied (4), from the intensity value (s) detected. 25. Optical method according to the preceding claim characterized in that it further comprises the following steps: - creation from the beam of polarized light of a reference beam not passing through the mobile polarizer (3);

- detection of the intensity of the reference beam. 26. Optical method according to the preceding claim characterized in that it further comprises the following steps:

- modulation of the light intensity of the linearly polarized light beam;

- determination of the absolute angular position of the object to be studied (4) from:. of the calculation of the value of the real transfer function of the polarizer (3) corresponding to the slope of the straight line defined by at least two pairs of detected intensity values, couples formed by the intensity of a measurement beam and of the intensity of the reference beam, and. a transfer table grouping together the slope / corresponding angular position pairs.