EP1592960A1

EP1592960A1 - Device and method for measuring the proportion of a constituent in a liquid

Info

Publication number: EP1592960A1
Application number: EP03809350A
Authority: EP
Inventors: Floréal FRANCH
Original assignee: A P I Electronique S A; Api Electronique Sa
Current assignee: Api Elec
Priority date: 2002-10-23
Filing date: 2003-10-23
Publication date: 2005-11-09
Also published as: WO2004038389A1; FR2846419A1; FR2846419B1; AU2003301566A1

Abstract

The invention concerns a device for measuring the proportion of a constituent in a liquid, comprising: a first light source (LED1 102) emitting rays having at least one wavelength, towards a container (101) holding a volume of said liquid, at a predetermined dispersion angle; means for optical separation (120) of the container having an optical interface between the liquid and a solid, said optical interface receiving rays derived from said first light source, with angles of incidence close to a limiting angle of refraction corresponding to at least one wavelength emitted by said first light source for a predetermined proportion of said constituent, said optical interface partly reflecting said rays, with a reflection rate depending on the angle of incidence of said rays and on the proportion of said constituent in said liquid; a first sensor (CCD 104) which picks up either rays reflected or rays refracted by said optical interface, and which transmits signals representing picked up rays, and first means for analyzing (109) the signals transmitted by the first sensor delivering the proportion of constituent in said liquid, based on the signals transmitted by said first sensor.

Description

DEVICE AND METHOD FOR MEASURING THE RATE OF A COMPONENT IN A LIQUID

The present invention relates to a device and a method for measuring the level of a component in a liquid. It applies, in particular, to the measurement of the sugar level in fermenting musts. Current sugar measurement devices impose complex, slow and expensive operations. For example, density measurements have traditionally been carried out for decades using a mustimeter (plunger density meter with a built-in mercury thermometer). They allow you to know the progress of the fermentation. However, its use is tedious and subject to many reading errors. The results obtained are recorded manually to draw the curve of variation of the density and the temperature as a function of time.

The present invention intends to remedy these drawbacks.

To this end, the present invention relates to a device for measuring the rate of a component in a liquid, characterized in that it comprises: - a first light source emitting rays having at least one wavelength, in the direction d 'a container containing a volume of said liquid, in a predetermined dispersion angle;

a means of optical separation of the container having an optical interface between the liquid and a solid, said optical interface receiving rays from said first light source, with angles of incidence close to a limiting angle of refraction corresponding to at least a wavelength emitted by said first light source for a predetermined rate of said component, said optical interface partially reflecting said rays, with a reflection rate depending on the angle of incidence of said rays and the rate of said component in said liquid; a first sensor which picks up either reflected rays or rays refracted by said optical interface, and which transmits signals representative of the rays picked up, and

a first means of analysis of the signals transmitted by the first sensor providing the level of component in said liquid, as a function of the signals transmitted by said first sensor.

Thanks to these provisions, the component level (for example of sugar or alcohol) is determined as a function of the reflection or refraction rates which itself depends on the optical index of the liquid which is a function of the component level in the liquid.

According to particular characteristics, the first means of analyzing the rays received is adapted to determine the position of a separator. Thanks to these provisions, the operation of the device is simple and effective, the separator corresponding to the limiting angle of refraction.

According to particular characteristics, the first light source is adapted to successively emit rays having different wavelengths. Thanks to these provisions, the refractive index of the liquid, which depends on the level of component to be measured in the liquid is determined for several wavelengths, which improves the accuracy of the measurement.

According to particular characteristics, the measurement device as succinctly described above further comprises: - a second light source emitting light rays having wavelengths different from the wavelengths of the rays emitted by the first source light,

- a second sensor which captures the light rays coming from the second source and reflected at a second interface between the liquid and a solid, - a third sensor which captures the light rays coming from the second source and refracted at said second interface and

a second means for analyzing the rays received providing said component rate as a function of the signals emitted by the second and third sensors.

Thanks to these provisions, the rate of a component is determined, possibly different from the component whose rate is determined by the first means of analysis, in the liquid and, optionally, the rate of one of its components is taken into account. to correct the measurement of another of these components. For example, the first means of analysis determines a sugar level and the second means of analysis determines an alcohol level.

According to particular characteristics, the first sensor is an image sensor. Thanks to these provisions, the distribution of the reflected or refracted light rays can be treated by the first means of analysis.

According to particular characteristics, the first analysis means determines a slope of the distribution of the light rays received by said first image sensor. Thanks to these provisions, the processing carried out by the first analysis means is easy. According to particular characteristics, the device as succinctly described above comprises a multiplexer and the first light source comprises a plurality of light sources controlled in transmission multiplexed by said multiplexer. Thanks to these provisions, several wavelengths are used successively and the accuracy of the measurement is improved. According to particular characteristics, the device as succinctly described above comprises a temperature sensor. Thanks to these provisions, the corrective factors due to temperature variations can be corrected and / or the reliability of the measurement can be evaluated and / or the analysis means only works for a predetermined temperature range. The present invention also relates to a fermentation control device characterized in that it comprises a measurement device as succinctly set out above.

The present invention also relates to a portable device comprising an energy source, characterized in that it comprises a measurement device as succinctly set out above. The present invention also relates to a method for measuring the rate of a component in a liquid using an optical separation means having an optical interface between the liquid and a solid, characterized in that it comprises:

a step of emission by a first light source emitting rays having at least one wavelength, in the direction of a container containing a volume of said liquid, in a predetermined dispersion angle, said optical interface receiving, around a limiting angle of refraction corresponding to at least one of said wavelengths, the rays having passed through said container, said optical interface partially reflecting said rays, with a reflection rate dependent on the angle of incidence of said rays; a step of capturing the reflected rays and / or the refracted rays by said optical interface, and

- A step of analyzing the rays captured during said capture step to provide said component rate as a function of the reflection rates of said rays.

Other advantages, aims and characteristics of the present invention will emerge from the description which follows, given for explanatory purposes with reference to the appended drawings in which:

- Figures 1A and 1B schematically shows the measurement part of the device according to the present invention;

- Figure 2 shows a succession of measurement steps implementing the device illustrated in Figures 1A and 1B;

- Figure 3 shows an operating flow diagram of the device illustrated in Figures 1A and 1B;

- Figure 4 shows electronic circuits implemented in the device illustrated in Figures 1A and 1B; and - Figure 5 shows, schematically, an alcoholic fermentation control device implementing the process object of the present invention.

Before detailing the figures, the following are the scientific and technical foundations necessary for a good understanding of its technical characteristics. Refraction is the deflection of light rays passing obliquely from one transparent medium to another.

This deviation follows the laws of Snell-Descarte:

First law: “The incident ray, the normal and the refracted ray are in the same plane. "

Second law: “When the angle of incidence is varied, there is a constant relationship between the sines of the angles of incidence and of refraction. This constant relationship, however, depends on the two environments considered. "

The equations which follow from these laws are: sin (î) / sin (r) = n (constant called refractive index). n = n2 / n1 (ni and n2 being the indices with respect to the vacuum determined by comparing the speed of light in the medium considered and the speed of propagation of light in vacuum).

The general equation of refraction is therefore: ni .sin (î) = n2.sin (r) In accordance with these laws, the deviation of the light ray varies in the same direction as the angle of incidence: plus the angle the greater the incidence, the greater the deviation of the light ray. For light rays incident on an interface between a first medium having a first refractive index and a second medium having a second refractive index less than the first index, there is an angle of incidence from which the reflection is total (all the rays are reflected). This angle is called "limit refraction angle". It depends on the refractive indices of each medium.

The use of a source which emits a beam of light rays, having different angles of incidence at the separation of the two media around the limiting angle of refraction, therefore results in refracted rays, when the angle d incidence is less than the limiting angle of refraction and of the reflected rays, when the angle of incidence is greater than the limiting angle of refraction (reflection rate of 100%) and when the angle of incidence is smaller at the refractive limit angle (reflection rate less than 100%). This is reflected, around the direction corresponding to the limiting angle of refraction, by the rays reflected at 100%, on the one hand and, by a more obscure zone, corresponding to the rays reflected at a rate lower than 100%. The position of the limit of these two zones, called "separator", is in direct relation with the ratio of the refractive indices of each of the media. If the index of one of the media is known and fixed, and that of the other unknown, the determination of the position of the separator makes it possible to measure the index of refraction. Indeed, the position of the separator corresponds to equality: n1.sin (î) = nx.sin (r) hence nx = n1.sin (î) / sin (r)

With: nx unknown refractive index sin (î) defined by the position of the source, sin (r) measured by the position of the separator, and ni refractive index of the known medium.

The present invention uses this principle to measure the level of a component in a liquid, when this rate varies the refractive index of the liquid (see LED1 102 in FIG. 1A and 1B).

In the frequency fields of the radiation for which the absorption by the unknown medium is small, the modifications of the propagation speed of a radiation, which correspond to the modifications of index, thus provide information on the nature of the medium crossed. .

For a constant temperature, the absolute refractive index of a medium relative to the vacuum is defined as the ratio clv of the propagation speed c of a radiation monochromatic electromagnetic in the vacuum at its speed v in this medium. The same liquid therefore has different indices depending on the wavelengths used for the measurement. Thus in a complex liquid, the various indices measured at different wavelengths make it possible to evaluate the rate of a set of known components compared to the set of all components.

In particular embodiments of the present invention, different wavelengths are used to determine the level of at least one component in a liquid, for example a level of sugar and / or alcohol in a must. fermentation (see multiled LED1 102 in Figure 1A and 1B). It is recalled, with regard to the absorption of an electromagnetic wave, that the radiation absorbed by a sample makes it possible to identify the transitions between the energy levels and to deduce therefrom information on the structure of the molecule. In the near infrared, the energies involved are mainly vibrational energies. The infrared absorption, by the molecule, brings into play two types of vibration: - the vibrations of elongations corresponding to the stretching of a bond AB denoted v _AB) and

- the vibrations of deformation (or bending) corresponding to the variation of a valence angle, noted δ _AB

We can model the molecule as a set of oscillators linked to each other, mechanically coupled. The excitation by infrared of one of the oscillators results in the simultaneous excitation of all the oscillators, both in elongation and in deformation. All the atoms of the molecule therefore vibrate with the same frequency v around their equilibrium position. Each of these whole molecule vibrations is called the normal mode of vibration.

For a given molecule, the deformation vibrations have lower frequencies than the excitation vibrations and therefore correspond to lower energies.

For the mechanical oscillator to be excited and to enter into vibration, there must be an interaction with the electromagnetic wave leading to an energy transfer. This occurs when the vibration varies the dipole moment of the molecule and therefore the interaction between the dipole and the electric field of the wave. The energy of vibration being quantified, the transition will occur when we have h. v = ΔE, energy difference between the fundamental level and the first level of excited vibration. The consequence is that only the vibrations varying the dipole moment are active in infrared and that the absorbed intensity is all the greater the greater this variation. The strongest absorptions are often due to very polar groups such as C = 0, N-O, O-H, etc. These latter groups have bonds which can be excited almost independently of the rest of the molecule. They will therefore give rise to characteristic absorption frequencies which depend little on their environment.

The experiments carried out by the inventors on this principle in infrared spectroscopy have thus shown that for the group v ₀ -H (bond between a hydrogen atom and a oxygen atom) the absorption of the alcohol molecule is maximum in the band from 3300 to 3500 nanometers, that is to say in the near infrared.

In particular embodiments of the present invention, this principle is used to measure the level of a component in a liquid (see LED2 103 in FIG. 1A): the measurement of the refractive index using a particular wavelength for which the absorption is maximum for a given component, makes it possible to evaluate the influence of this component on the measurement. Calculation and correlation algorithms then give the actual concentration of each component whose concentration we want to know.

As illustrated in FIGS. 1A and 1B, the measurement system 100 according to the present invention (hereinafter called "sugar sensor") consists of an optical part called "prism" 101, from a first light source to light-emitting diode LED1 102 and a second light source with light-emitting diode LED2 103, of a linear image sensor with charge transfer "CCD" 104 (for "charge coupled device, in French" transfer device of charges "or" DTC "), of two photodiodes PD1 105 and PD2 106, of a temperature sensor 107, of an electronic assembly 108 comprising a microprocessor 109 and of the circuits illustrated in FIG. 4 and implementing software for piloting and interpretation of measurements.

The prism101 is made of optical silica and consists of a cylindrical main body with four opposite faces two by two and bevelled at 70 degrees of angle, with respect to one of the sections of the cylinder called “measuring surface” 120. The liquid to be analyzed circulates in prism 101 which has walls of constant thickness.

On a pair of opposite and bevelled faces, are glued, on the light source side LED1 102, an optical waveguide 111 which standardizes the light emitted and, on the sensor side CCD104, a focusing lens 112. On the other couple are glued, near infrared source side LED2 103, a focusing lens 113 preceded by an optical filter 114 with a bandwidth of 3300 to 3500 nanometers and, on the photodiodes PD1 105 and PD2 106 side, a focusing lens 115 (respectively 125) followed also of an optical filter 116 (respectively 126) having the same bandwidth as the optical filter 114.

As a variant (not shown), on the periphery of the cylinder which constitutes the optical part, other pairs of light sources / sensors (LED and photodiode) are distributed similar to the pair LED2 103 / photodiodes PD1 105 and PD2 106, within the dimensional limit various elements constituting the whole.

In the particular embodiment illustrated in FIGS. 1A and 1B, the light source LED1 102 consists of a multi-frequency (or "multiled") LED capable of alternately emitting light rays having several wavelengths in the fields visible and not visible, with, for each of them, a spectral band of width from 10 to 20 nanometers. The wavelength of 589.3 nanometers (wavelength of the sodium line usually used for sugar level measurements in traditional refractometry) is taken as a reference. The use of a scan of several wavelengths is carried out to measuring the sugar level of any medium, and more particularly for measuring the sugar level in complex liquids such as grape must in fermentation.

The light-emitting diode LED2 103 emits light rays in the near infrared range in order to determine the alcohol level in the liquid analyzed. With regard to the measurement of the sugar level carried out using the LED1 102 light source, the optical measurement is based on the measurement of the refractive index of the liquid in contact with the measurement surface 120. This index defines a limit angle of total reflection on this surface. The value of this angle varies from 63 ° to 70 °, or an amplitude of variation of 7 °, for a range of refractive index of the liquid going from 1.3329 to 1.4000 for a wavelength of 589, 3 pressure gauges at 20 ° C. The choice of the bevel angle of 70 ° of the prism 101 corresponds to these limiting refraction angle values. In general, if the limit angle varies, in the component rate measurement range, from A ° and B °, B being greater than A, the bevel angle will be chosen around the value of B ° .

The focusing lens 112 delimits the projection of the 7 ° of angular variation of the total reflection so as to cover the measurement length of the CCD sensor 104.

The measurements made thanks to the implementation of the second light source LED2 103 and of the photodiodes PD1 105 and PD2 106, on another axis of the prism 101 are intended to measure the level of alcohol contained in the unknown medium: the set constituted by the light source LED2 103 and the diodes PD1 105 and PD2 106 operate on an axis of the prism different from that used previously. It is observed that the same axis could also have been used since the wavelengths of the light rays received by the sensors 104, 105 and 106 do not overlap. However, this solution would have posed unnecessary mechanical problems.

The optical waveguide 111 is bonded to the optical unit opposite the light source LED2 103 followed, itself, by the selective optical filter 114 intended to keep only the band of 3300 to 3500 nanometers of the spectrum. A focusing lens 113 is placed on the opposite face opposite the photodiode PD2 106, this lens also being followed by a filter 124 identical to the filter 114. The angles of the waveguide 111 and of the focusing lens 113 are calculated so as to have, in the near infrared band used, a range of refractive indices which covers the range of variation of an alcohol concentration from 0% by volume up to 25% by volume. The photodiode PD1 105 measures the light resulting from the total reflection on the measurement surface 120 while the diode PD2 106 measures, at the source, the intensity of the light projected by the focusing lens 113.

The difference between these two measurements allows us to know the absorbance due to the unknown medium. The latter corresponds to the refractive index for the near infrared band of 3300 to 3500 nanometers at the wavelength used and to the energy used for the vibration of the atoms constituting the alcohol molecules. Case of alcoholic fermentation: A measurement, with the measuring device illustrated in Figures 1A and 1B, is initially carried out on the must before fermentation. It is used as a reference for the following measurements carried out during alcoholic fermentation. Measurements in the near infrared wavelength band (emitted by the second light source

LED2 103) allow to evaluate the influence of alcohol on the refractive index and therefore to introduce a corrective factor in the analysis of the refractive index obtained by the use of rays from the first light source LED1 102) which provides the actual concentration of sugars.

The calculated refractive indices are converted to sugar concentration and to BRIX

(official unit of measure representing the ratio of sugar in grams to 100 grams of aqueous solution). Additional conversion tables give an approximate value of the density of sugar in the must.

A digital conversion is directly performed at the output of the CCD sensor 104, in it. The currents of the photodiodes PD1 105 and PD2 106 are directly applied to the multiplexed input of the analog digital converter of the microprocessor 109 of the electronics 108. The processing by the microprocessor 109 makes it possible to drive in pulse the light-emitting sources LED1 102 and LED2 103 in synchronism with the CCD sensor 104 and the photodiodes PD1 105 and PD2 106.

For the CCD 104, the measurement of the position of the separator (separation of a light area and a darker area) is carried out for each pulse of the first light-emitting source LED1 102. In the embodiment illustrated in FIGS. 1A and 1 B, for each cycle, five ray pulses having five different wavelengths are emitted by the first light source LED1 102. The focusing lens delimits the projection of the 7 ° angular variation of the total reflection so as to cover the CCD 104 measurement length. All the values thus obtained are then corrected in temperature and transmitted to a processing unit or central unit (not shown). This can be included in the body of the sensor or remote. In the latter case, the information is transmitted by an encrypted digital link (see Figure 4). The transmission can be wired or wireless (radio, optical or ultrasonic transmission). In the case of a processing unit included in the measurement system 100, this consists of a second structure comprising a microprocessor, different from the microprocessor 109 (but which, as a variant, may be the microprocessor 109), and implementing the algorithms and libraries necessary for the correlation between optical measurements and sugar and alcohol levels. The microprocessor 109 generates on its outputs P1, P2, P3, P4 and P5 pulse width modulated pulse trains (in English "PWM" for "Puise Width Modulation") which make the Mos 130 transistors respectively connected to these outputs. For each Mos 130 transistor, an RC 131 integration circuit smoothes the current thus generated by the Mos 130 transistors. Each current applied to the terminals of the first light source LED1 102 (in fact, here a multiled source) generates pulses of light for each of light-emitting diodes that make up the first light source LED1 102. The light wavelength emitted by the first light source LED1 102 depends on the respective values of the currents applied to each of the light-emitting diodes that make up the multiled LED1 102. Likewise, the intensity light in the near infrared emitted by the source LED2 103 depends on the frequency and the width of the current pulses applied to it. The sequence of a complete measurement is illustrated in Figure 2 and includes:

for a first wavelength emitted by the first light source LED1 102, a step 201 for transmitting control pulses by the microprocessor 109, a step for reading 202 from the sensor 104, a step for determining the splitter 203, a step 204 of transmitting the result of step 203 to the central unit;

for a second wavelength emitted by the first light source LED1 102, a step 211 for transmitting control pulses by the microprocessor 109, a step for reading 212 from the sensor 104, a step for determining the splitter 213, a step of transmitting 214 the result of step 203 to the central unit;

- And so on for all the wavelengths that the first light source is capable of emitting; and

- For the second light source LED2 103, a step 231 of transmitting control pulses by the microprocessor 109, a step of reading 232 of the photodiode PD1 105, a step of reading 233 of the photodiode PD2 106 simultaneously with the step 232, a step 234 of determining the difference between the values read during steps 232 and 233, and a step 235 of transmitting the result of step 234 to the central unit.

The search for the position of each separator, step 203, 213, ... is carried out by looking for the slope of the curve detected by the CCD 104 in accordance with the flow diagram illustrated in FIG. 3, which implements the algorithm next :

Let "x" be the pixel value of the position of the separator sought. During a step 301, a first search at low resolution (8 bits) of the separator is carried out: if

Σ x + 8 <-—- x - ^

VAL ^- _L_ VAL ^ ^> 8 bits * ⁺¹ * - *

The value of x which satisfies the inequality is Xi and is the intermediate position sought. If no value x satisfies this equation, a sensor fault message is generated, step 302.

Otherwise, during a step 303, a second search is carried out at 16 bit resolution for precise determination of the position of the separator: if

Xl + 15 _τ - χι

VAL - 2_, VAL 16 bits

Xi + 1 XI-15 The value of x which satisfies the inequality is Xs and is the position of the separator sought. Sβbits and S-iβbits are the values of the slopes for which the detection of the separator is validated.

The value of Sβbits is parameterized when characterizing the CCD sensor 104 and the value Siβbits is parameterized according to the zone in which the position of the separator is determined determined during step 303. The zones are defined according to the resolution of the CCD 104 during a step 304:

- a zone corresponds to high concentrations resulting in a position of the separator on a first part of the useful zone of the CCD 104. - a zone corresponds to low concentrations resulting in a position of the separator on a second part of the useful zone of the CCD 104.

These zones, as well as the corresponding coefficients, are determined for each CCD sensor 04 during the sensor characterization phase and are stored in non-volatile Flash memory 410 by the microprocessor 109 or by the CPU 400 (FIG. 4). From the result of the last equation calculated for a 16-bit resolution, a value of the slope is calculated, step 305, by dividing the number of pixels used: let PL be the limit value of the slope for a correct measurement

VAL ^- -> PL

32

If the slope value found is greater than the limit value PL (limit slope), the measurement is validated, without any particular comment, step 308. If the slope value found is less than the limit value PL and a separator value is still found, the measurement is accompanied by a comment indicating the likely fouling state of the sensor (setting an alarm bit to 1), step 306.

In the case where no separator is detected, a message is generated (setting to 1 of a sensor fault bit), step 307.

The step sequence illustrated in FIG. 3 is applied for each value of wavelength generated by the source LED1 102. It is therefore N values of separator positions each associated with a different wavelength which are transmitted to the 'calculation unit (UC). To ensure better accuracy in the case of liquids having gradients of rate variations measurable by the sensor (case of fermenting musts), a series of several measurements is carried out for each wavelength and an average value of the separator is calculated on the basis of the results of the various measurements. The measurements taken into account for this average can be taken one after the other for each wavelength or extracted from several measurement cycles of several wavelengths. The values of the positions of the separators, the absorbance and the temperature are transmitted to the central unit CPU 400 (see FIG. 4) in the form of coded digital frames; either directly in the case where it is physically in the sensor housing; either via a wire, radio frequency, optical or ultrasonic link. We observe, in Figure 4, the central processing unit or CPU 400, associated with a reception interface 401 (line receiver, radio frequency, optical and ultrasonic receiver and demodulator), with a decryption module 402, with a memory program storage 409, a result storage memory 410, a local or remote display screen 411 and a digital link to a communication interface 412. The frames decoded by the decryption module 402 are processed by the CPU 400 to extract the desired sugar level value.

The calculation method is based on the correlation between the value found for each wavelength with tables of reducing sugar and alcohol concentrations determined experimentally on musts at several stages of fermentation. Additional tables giving the concentrations of any other component involved in the fermentation process can be added (fructose, glucose ...).

The experimental values were carried out on musts from a selection of several grape varieties representative of the average wine production in the main countries of the world producing quality wines.

The measurements of reducing sugars (total, fructose and glucose) and alcohol levels are carried out in the oenology laboratory by the enzymatic assay method with a continuous spectrum photometer in double samples. A correlation with measurements carried out in nuclear magnetic resonance (NMR) completed the results obtained.

All the sugar concentration measurements thus carried out were converted into BRIX (official unit of measurement representing the ratio of the quantity of sugar, in grams, to 100 grams of aqueous solution).

The measurement tables thus generated are therefore of the form:

With types of white, rosé or red wine making. The potential alcohol is the presumed value of the volume percentage of alcohol at the end of fermentation given for a transformation of 16.83 g of sugar (this value can be changed depending on the type of yeast) into 1% alcohol. The sugar level measurement is carried out before fermentation. It is essential for characterizing the initial data and choosing the corresponding table in the database.

With e = number of samples of eo e _m (e0 measurement of the wort before fermentation) λ = wavelength of each measurement of the first to the n measurement ^me

The experimental tables thus obtained are stored in memory 409 or 410. For each wavelength, each of the measurements made by the CCD sensor 104 is compared with the values of the sugar concentrations, density and alcohol levels in the various tables. which correspond to the type of vinification in progress (red, white or rosé) as well as to the selected grape variety. A temperature correction is performed on each of the measurements to make the comparisons at identical temperature. A calculation of the standard deviation on all the measurements falling within a template is carried out with a coefficient of variation of 5%.

The results for sugar concentration and density are the averages of the various results falling within this template. The result from this calculation is then compared to the last measurement carried out using a likelihood algorithm which calculates the theoretical transformation of sugar into alcohol corrected by the real variations of the previous measurements and deduces from it the standard deviation between the theoretical value and the value found by the measurement. The result is compared to a template for validation.

The final values of the sugar level and density are then validated or corrected by this last algorithm.

The prism 101, in which liquid circulates from the fermentation medium, performs the measurements in real time at the rate defined by the user. The measurement results can be displayed locally (screen 411) and / or be transmitted (interface 412):

- to a centralized IT unit, by means of digital frames supported by an industrial field bus to which all the sensors of an installation are connected;

- to a display or to a regulator by an analog link for each physico-chemical quantity measured (current loop 4-20 mA for example).

The fermentation process is traditionally carried out by maintaining a temperature range in which the yeasts can have sufficient activity to transform the sugar into alcohol (thermoregulation). A density measurement check using a mustimeter (performed one or more times a day) allows the temperature range to be corrected if necessary. Numerous fermentation incidents due to runaway yeast activity, which have resulted in a degradation of the process and therefore of the quality of the final product, have shown the limits of this type of regulation. It has been shown that within the limits of a given temperature range, it is sufficient to increase or decrease the temperature to favor or disadvantage the multiplication of yeasts. An application derived from the process of the invention: the management of alcoholic fermentation. Starting from the principles set out above, the sugar level sensor 100 also makes it possible to regulate the fermentation process by measuring the sugar level by acting on the temperature. The operation is based on the measurement of the temperature and the sugar level carried out by the sugar level sensor 100. This is permanently immersed in the fermenting liquid 502 (wort for example). It is hung by its electrical connection cable and connected to a communication interface via an industrial digital bus which allows dialogue between several sugar sensors 100 and a central control unit 500 (see Figure 5). A fermentation curve giving the quantity of sugar transformed into alcohol per unit of time is established by the user. This curve can have as many slopes (transformation gradient) as necessary. It serves as a deposit for the management of the industrial fermentation system. Maximum deviations are defined in relation to this curve. A regulation algorithm (fuzzy logic, PID ...) actuates valves 501 which control the passage of heat transfer fluids so as to cool or warm the fermentation medium 502 to slow down or activate the activity of yeasts which carry out the transformation of the sugar in alcohol. The heating or cooling action is triggered so as to keep the sugar level within the limit of the template defined by the setpoint curve.

The block diagram of the operation of such a system is given in FIG. 5. Such a system can be integrated into a global computerized process which allows, with the same central unit, to control the whole of a cellar made up of several tanks.

Application of the invention to a point measurement system: the refractometric rod (not shown). During alcoholic fermentation, spot measurements of sugar and temperature can be made (for example, twice a day). The refractometric cane gives the sugar level directly.

Based on the measurement principle object of the present invention, it consists of the sensor 100 in the lower part, fixed to a handle (for monitoring fermentations in solid phase, in the case of wine making in red) or to a cable (for fermentations in the liquid phase, in the case of white or rosé wine making) of sufficient length to allow the wort to be reached in fermentation and in the upper part with, incorporated in this handle or connected to the cable, a set of batteries with a radio transmitter. The digital frames are transmitted to a portable terminal including a receiver. The whole is completely autonomous. The terminal records the sugar level and temperature measurements made on all the tanks with the time stamp and operator identification. The files thus created can be transferred to a computer system for plotting curves and storage.

This refractometric rod therefore makes it possible to combine measurement precision, ease of implementation and traceability.

It is observed that what has been implemented in the particular embodiments illustrated in FIGS. 1 to 5 use reflection to determine the position of the separator. In other embodiments not shown, the separator is determined by capturing the light rays refracted on the measurement surface 120. These two modes of determining the separator can be used, alone or in combination, in a device according to the present invention.

Likewise, the present invention does not only apply to the determination of the level of components in a liquid but can be applied to any gaseous or solid medium having a non-zero transparency for at least one wavelength. In these cases, it is preferable that the product to be analyzed has a plane or known optical interface over at least part of its periphery. For example, the combined analysis, possibly outside the visible spectrum, of the spectrum of incident solar rays and refracted rays reflected by drops of water, known in the phenomenon of rainbows, would, in accordance with the present invention, to determine components entering atmospheres, for example polluted. Likewise, the present invention can be applied to determining, in real time, the concentrations and the components diluted in the water of a swimming pool, such as chlorine, or the components entering into a glass or a crystal.

Claims

1 - Device for measuring the rate of a component in a liquid, characterized in that it comprises: - a first light source (LED1 102) emitting rays having at least one wavelength, in the direction of a container (101) containing a volume of said liquid, in a predetermined dispersion angle;

- an optical separation means (120) of the container having an optical interface between the liquid and a solid, said optical interface receiving rays from said first light source, with angles of incidence close to a corresponding angle of refraction at least one wavelength emitted by said first light source for a predetermined rate of said component, said optical interface partially reflecting said rays, with a reflection rate depending on the angle of incidence of said rays and the rate of said component in said liquid; a first sensor (CCD 104) which picks up either reflected rays or rays refracted by said optical interface, and which transmits signals representative of the rays picked up, and

- a first means of analysis (109, 400) of the signals transmitted by the first sensor providing the level of component in said liquid, as a function of the signals transmitted by said first sensor.

2 - Device according to claim 1, characterized in that the first means for analyzing the captured rays (109, 400) is adapted to determine the position of a separator.

3 - Device according to any one of claims 1 or 2, characterized in that the first light source (LED1 102) is adapted to successively emit rays having different wavelengths.

4 - Device according to any one of claims 1 to 3, characterized in that it further comprises:

a second light source (LED2 103) emitting light rays having wavelengths different from the wavelengths of the rays emitted by the first light source (LED1 102),

- a second sensor (PD2 106) which captures the light rays coming from the second source and reflected at a second interface between the liquid and a solid,

- a third sensor (PD1 105) which captures the light rays coming from the second source and refracted at said second interface and - a second means of analysis of the captured rays providing said component rate as a function of the signals emitted by the second and third sensors.

5 - Device according to any one of claims 1 to 4, characterized in that the first sensor (CCD 104) is an image sensor. 6 - Device according to claim 5, characterized in that the first analysis means (109, 400) determines a slope of the distribution of the light rays received by said first sensor (CCD 104).

7 - Device according to any one of claims 1 to 6, characterized in that it comprises a multiplexer (109, 130, 131) and in that the first light source (LED1 102) comprises a plurality of light sources controlled by transmission multiplexed by said multiplexer.

8 - Device according to any one of claims 1 to 7, characterized in that it comprises a temperature sensor (102). 9 - Device for piloting fermentation, characterized in that it comprises a measuring device according to any one of Claims 1 to 8.

10 - Portable device comprising an energy source, characterized in that it comprises a measuring device according to any one of claims 1 to 8.

11 - Method for measuring the rate of a component in a liquid using an optical separation means having an optical interface between the liquid and a solid, characterized in that it comprises:

a step of emission by a first light source emitting rays having at least one wavelength, in the direction of a container containing a volume of said liquid, in a predetermined dispersion angle, said optical interface receiving, around a limiting angle of refraction corresponding to at least one of said wavelengths, the rays having passed through said container, said optical interface partially reflecting said rays, with a reflection rate dependent on the angle of incidence of said rays;

a step of capturing the reflected rays and / or the rays refracted by said optical interface, and a step of analyzing the rays picked up during said capturing step in order to provide said component rate as a function of the reflection rates of said rays .