EP4305413A1

EP4305413A1 - Method for characterizing an analyte present in a gas sample containing at least one parasitic chemical species

Info

Publication number: EP4305413A1
Application number: EP22709746.6A
Authority: EP
Inventors: Pierre MAHO
Original assignee: Aryballe SA
Current assignee: Aryballe SA
Priority date: 2021-03-08
Filing date: 2022-03-04
Publication date: 2024-01-17
Also published as: CN117098992A; KR20230155542A; FR3120445B1; WO2022189292A1; FR3120445A1

Abstract

The invention relates to a method for characterizing an analyte A present in a gas sample using an electronic nose comprising M sensitive sites, a parasitic chemical species P being present in the gas sample, comprising: o a phase 100 of acquiring N first signatures, where N>1, of the gas samples containing the analyte A and the parasitic species P, the gas samples exhibiting deviations Δc _p(n) which differ from one gas sample to the next; o a phase 200 of solving an optimization problem so as to obtain N corrected signatures, characterizing the analyte A present in the N gas samples, from the N first signatures, by optimizing two objective functions.

Description

METHOD FOR CHARACTERIZING AN ANALYTE PRESENT IN A GAS SAMPLE CONTAINING AT LEAST ONE PARASITE CHEMICAL SPECIES

TECHNICAL AREA

[001] The field of the invention is that of the characterization of an analyte present in a gaseous sample by means of an electronic nose. The gaseous sample here comprises, in addition to the analyte to be characterized, at least one chemical species capable of also interacting with the functionalized measurement surface of the electronic nose.

PRIOR ART

[002] The ability to characterize analytes contained in gaseous samples, for example odorous molecules or volatile organic compounds, is an increasingly important issue, particularly in the fields of health, the food industry, the perfumery industry (scents), olfactory comfort in confined public or private places (cars, hotels, shared places, etc.), etc. The characterization of such analytes present in a gaseous sample can be carried out by a characterization system called "electronic nose".

[003] Different characterization approaches exist, which differ from one another in particular by the need or not to have to “mark” the analytes or the receptors beforehand with a revealing agent. Unlike, for example, measurement by fluorescence which requires the use of such markers, measurement using surface plasmon resonance (SPR for Surface Plasmon Resonance, in English), and measurement using an interferometric principle of the Mach- Zehnder (MZI, for Mach-Zehnder Interferometer, in English), are so-called label-free techniques.

[004] In an electronic nose using SPR or MZI technology, the analyte present in a gaseous sample interacts by adsorption/desorption with receptors located in several distinct sensitive sites of a functionalized measurement surface. It is then a question of detecting in real time a measurement signal associated with each of the sensitive sites, representative of the adsorption/desorption interactions between the analyte and the receptors, in response to a primary signal. The measurement signals can be optical signals representative of a temporal variation of the local refractive index due to the interactions of the analyte with the receptors. In the context of SPR technology, for example, the intensity of optical signals is measured in real time. coming from the different sensitive sites, these optical signals being a reflected part of a primary optical signal emitted by a light source. The intensity of each optical signal detected by an optical sensor is directly correlated to the adsorption/desorption interactions of the analyte with the receptors. Application WO2018/158458 describes an example of such an electronic nose in SPR technology.

[005] Insofar as the chemical and/or physical affinity of interaction of the analyte with the receptors is not known, the characterization of the analyte then amounts to determining a value or a variation of a parameter representative of the adsorption/desorption interactions of the analyte with the receptors, here representative of the temporal variation of the local refractive index for each of the sensitive sites. An interaction motif, or a signature, which characterizes the analyte is thus obtained. Indeed, the adsorption/desorption interactions of the analyte on sensitive sites (functionalized surfaces) benefiting from differentiated adsorption characteristics make it possible to account for the molecules present in the gas which have been attached to the surface of the different sensitive sites. .

[006] Figures IA and IB illustrate an example of an electronic nose of the SPR type as described in document WO2018/158458. This type of electronic nose 1 generally comprises a fluid supply device 2, a characterization device 3 by SPR imaging, and a processing unit (not shown).

[007] The characterization device 3 comprises a measurement chamber 4 intended to receive the gaseous sample, in which is located a measurement surface 5 on which is located a matrix of sensitive sites. The measurement surface 5 is formed of a metallic layer on which are fixed various receptors suitable for interacting with the analyte, the receptors being arranged so as to form various sensitive sites distinct from one another. These receivers are then located at the interface between the metallic layer and a dielectric medium, here a gaseous medium.

[008] This characterization device 3 further comprises a light source 7 of a primary optical signal and an image sensor 8. The light source 7 is adapted to emit the primary optical signal in the direction of the measurement surface 5, at a working angle 0 _R making it possible to generate surface plasmons therein. The reflected part of the primary optical signal, forming an optical measurement signal, is then detected by the image sensor 8. The intensity of the optical measurement signal depends locally on the refractive index of the measurement surface 5, which itself depends on the surface plasmons generated and the quantity of matter located at the level of each sensitive site, this quantity of matter varying over time according to the interactions between the analyte and the receptors. The measurement surface 5 comprises a plurality of 6 _m sensitive sites, here M distinct sensitive sites, functionalized by the presence of receptors with which the analyte to be characterized can interact by adsorption/desorption.

[009] The electronic nose processing unit is suitable for analyzing the "sensorgrams", that is to say the signals corresponding to the temporal evolution of the parameter representative of the adsorption/desorption interactions of the analyte with the receptors of each of the various sensitive sites 6 _m , with the aim of extracting therefrom information on the kinetics of interaction (adsorption and desorption) of the analyte with the receptors. These sensorgrams can be so-called useful signals Su _m (t) corresponding to the temporal evolution of the variation A%R _m (t) of the reflectivity associated with each of the sensitive sites 6 _m . The reflectivity %R is here the ratio between the intensity of the optical measurement signal detected by the image sensor 8 over the intensity of the primary optical signal emitted by the light source 7. The variation in reflectivity A%R is obtained by subtracting from the temporal evolution of the reflectivity %R(t) a reference value (baseline, in English) associated with the only gas present inside the measurement chamber 4, independently of the analyte.

[0010] Finally, the fluid supply device 2 is adapted to introduce the analyte into the measurement chamber 4 under conditions allowing the analysis of the sensorgrams and therefore the characterization of the analyte. As such, the article by Brenet et al. entitled Highly-Selective Optoelectronic Nose based on Surface Plasmon Resonance Imaging for Sensing Gas Phase Volatile Organic Compounds, Anal. Chem. 2018, 90, 16, 9879-9887, describes a method for characterizing a gaseous sample using an electronic nose of the SPR imaging type. This characterization method consists in supplying the measurement chamber with a gaseous sample in such a way that the kinetics of interaction between the analyte and the receptors reaches a steady state of equilibrium.

As such, Figure IC illustrates an example of sensorgrams Su _m (t) obtained by the electronic nose of fig.lA. The sensorgrams Su _m (t) correspond here to the temporal evolution of the variation of the reflectivity A%R _m (t) associated with each of the sensitive sites 6 _m . Thus, the characterization method comprises several successive fluidic injection steps, namely: a first reference phase Phi, in which a carrier gas alone, without the analyte, is brought into contact with the measurement surface. This carrier gas is generally identical to that of the gaseous sample; a second supply phase Ph2, in which the gaseous sample, formed from the carrier gas and the analyte to be characterized, is brought into contact with the measurement surface; and a third purge phase Ph3, in which the reference gas alone is again injected into the measurement chamber, so as to dissociate the analyte from the receptors, and to evacuate it from the measurement chamber. measure. The initial phase Phi makes it possible to acquire the reference value {baseline), mentioned above, which is then intended to be subtracted from the measurement signals S _m (t) to obtain useful signals Su _m (t) ( in other words the temporal evolution of the variation in reflectivity A%R _m (t) for each sensitive site of rank m). As indicated previously, this phase of fluidic injection is carried out so that the sensorgrams highlight the presence of a transitory regime of assimilation followed by a stationary regime of equilibrium. When this steady state of equilibrium is reached, the (steady state) equilibrium values Su _m,f of the useful signals Su _m (t) are extracted by the processing unit, and define the signature of the analyte.

[0013] However, it appears that the presence of other chemical species in the carrier gas, such as water molecules (relative humidity), can have an impact on the intensity of the optical measurement signal, as indicated the article by Shao et al. titled Mechanism and Characteristics ofHumidity Sensing with Polyvinyl Alcohol-Coated Fi ber Surface Plasmon Resonance Sensor, Sensors 2018, 18, 2029. In this article, the authors use an SPR sensor as a humidity sensor. However, in the context of an analyte characterization method using an electronic nose, the variation in relative humidity in the measurement chamber forms a measurement bias that degrades the quality of the characterization. Moreover, in the case where the relative humidity varies over long times, and therefore varies from one characterization to another for the same analyte and the same operating conditions, this induces a temporal drift which makes the signatures of the same analyte different from each other. [0014] Furthermore, document WO 2020/141281 A1 describes a method for characterizing target compounds, using an electronic nose type analysis system.

DISCLOSURE OF THE INVENTION

[0015] The object of the invention is to remedy at least in part the drawbacks of the prior art, and more particularly to propose a method for characterizing an analyte which makes it possible to limit or even eliminate the measurement noise associated with a variation in the concentration of one or more parasitic chemical species present between the initial phase Phi and the characterization phase Ph2 of fluidic injection. The characterization method thus makes it possible to improve the quality of characterization of the analyte. For this, the object of the invention is a method for characterizing an analyte A present in a gaseous sample located in contact with a measurement surface of an electronic nose, the measurement surface comprising M sites distinct from each other, of rank m ranging from 1 to M, having receptors adapted to interact by adsorption/desorption with the analyte A and with at least one so-called parasitic chemical species P present in the gaseous sample, the method comprising the following steps: fluidic injection, to bring into contact with the measurement surface: during a first phase Phi, a carrier gas which may contain the parasitic species P with a concentration cp _(n)i ; then, during a second phase Ph2, the gaseous sample containing the analyte A in concentration c _A(n) and the parasite species P in concentration c _P(n)f , the value c _P(n)f having a difference Acp _(n) not zero with the value c _P(n)i ; determination of a measurement signal S _(n,m) (t), during the fluidic injection step, representative of the interactions of the analyte A and of the parasite species P with the receptors, for each of the M sensitive sites; determination of first signatures Su _(n,m)f from the measurement signal S _(n,m) (te Ph2) associated with the gaseous sample, which having been corrected by a reference value S( _n , _m )i from the measurement signal S _(n,m) (te Phi) associated with the carrier gas; determination of a deviation Acp _(n) of concentration of the parasitic species P between the first phase Phi and the second phase Ph2; repetition of the previous steps, by incrementing the rank n until obtaining N first signatures representative of the interactions of the analyte A and the parasite species P with the receptors, with N>1, the gaseous samples being such that the Acp deviations _(n) are different in pairs; formation of a matrix of first signatures Su _AjP , of dimensions NxM, formed from the N first signatures Su _(n,m)f determined for the M sensitive sites; and of a vector of relative concentration Acp, of dimension N×1, from the N deviations Acp _(n) determined; determination of an estimated solution _| whose product forms a matrix of so-called corrected signatures Suc _A characterizing the analyte A present in the N gaseous samples, the estimated solution: minimizing the cost function _| maximizing the cost-function are defined variables as follows: oc _A is a concentration vector of analyte A, of dimension N, formed from the N concentration values c _A (n) of the gaseous samples, ok _A is an affinity vector of analyte A , of dimension M, formed of the M values of an affinity of interaction of the analyte A with the receptors of the sensitive sites, ok _P|A is an affinity vector of the parasite chemical species P, of dimension M, formed from the M values of an affinity of interaction of the parasitic chemical species P with the receptors of the sensitive sites in the presence of the analyte A. Some preferred but non-limiting aspects of this characterization method are as follows.

The step of determining the estimated solution can perform the minimization of the objective function f and the maximization of the objective function g at the same time, the values of the relative concentration vector all being positive.

The step of determining the estimated solution can be performed by an iterative algorithm, with iteration indicator i.

[0020] The step of determining the estimated solution may include a sub-step of determining the value of the variable taking into account the value c of the variable C _HAS and the value of the variable by a fixed point method.

[0021] The step of determining the estimated solution may include a sub-step of determining the value c _A ⁺¹⁾ of the variable C _A and the value of the variable k _A , given the value fc of the variable _{k P|A} having been determined, by a singular value decomposition of a matrix The step of determining the estimated solution may include a sub-step of minimizing the objective function f to obtain a plurality of Q local solutions minimizing the objective function f, with Q>1, followed by a maximization sub-step of the objective function g.

[0023] The minimization sub-step can provide Q local solutions formed each of the estimates of rank q going from 1 to Q, variables Q is here greater than 1: Q>1.

[0024] The maximization sub-step may comprise the determination of Q matrices of so-called corrected signatures

The method may comprise, following the determination of the Q corrected signature matrices Suc®, a normalization of each of the corrected signature matrices for obtain Q matrices of corrected and normalized signatures

The method may comprise, following the determination of the Q matrices of corrected and normalized signatures S, a determination of a variance score, for each of the Q matrices the variance score being defined as the trace of the covariance matrix for each of the Q matrices having a minimum score characterizing the analyte Now in the N gas samples. The method may comprise, following the determination of the Q matrices of corrected signatures S a determination of a norm of each of the Q matrices followed by a matrix identification having the maximum norm.

[0028] The matrix can be normalized, thus providing a matrix characterizing the analyte A present in the N gaseous samples.

The electronic nose may comprise a device for measuring interactions by adsorption/desorption of the optical type with surface plasmon resonance or of the Mach-Zehnder interferometry type.

[0030] the electronic nose may comprise a device for measuring interactions by adsorption/desorption of the resistive, piezoelectric, mechanical or acoustic type.

[0031] The electronic nose may include a fluid supply device suitable for carrying out the fluid injection step, a measuring device suitable for carrying out the step of determining the measurement signal, a sensor for measuring and determining the the relative concentration of the parasitic chemical species, and a processing unit adapted to implement the step of determining the estimated solution.

BRIEF DESCRIPTION OF DRAWINGS

Other aspects, objects, advantages and characteristics of the invention will appear better on reading the following detailed description of preferred embodiments thereof, given by way of non-limiting example, and made with reference in the appended drawings in which: FIG. 1A, already described, is a diagrammatic and partial view, in cross-section, of an electronic nose with SPR imaging according to an example of the prior art; FIG. 1B, already described, is a top view, schematic and partial, of a functionalized measurement surface of the electronic nose of FIG. 1A comprising M distinct sensitive sites; figure IC, already described, is an example of sensorgrams Su _m (t) obtained by the electronic nose of fig.lA, these sensorgrams corresponding here to the temporal evolution of the variation of the reflectivity A%R _m (t) associated with sensitive sites; FIG. 2 is a schematic and partial view of an electronic nose according to one embodiment; FIG. 3A is an example of three signatures obtained by a characterization method according to the prior art, highlighting the degradation of the characterization of the analyte due to a variation in concentration of a parasitic chemical species (here water in the vapor phase) between the initial phase Phi and the characterization phase Ph2; FIG. 3B illustrates the temporal evolution of the concentration cp of a parasitic chemical species, of the measurement signal S _m (t) and of the useful signal Su _m (t), in the case where there is no variation of the concentration c _P between the initial phase Phi and the characterization phase Ph2 (left part), and in the case where there is a non-zero variation in the concentration c _P between the initial phase Phi and the characterization phase Ph2 (right part); FIG. 4 is a flowchart of a characterization method according to a first embodiment; FIG. 5 is a flowchart of a characterization method according to a variant of the first embodiment; FIG. 6 is a flowchart of a characterization method according to a second embodiment; FIG. 7A illustrates examples of uncorrected signatures of analyte present in gaseous samples for which there has been a variation in the concentration c _P of the parasitic chemical species; Fig. 7B illustrates the analyte signatures of Fig. 7A corrected by the characterization method according to the second embodiment. DETAILED DISCUSSION OF PARTICULAR EMBODIMENTS

In the figures and in the following description, the same references represent identical or similar elements. In addition, the various elements are not shown to scale so as to favor the clarity of the figures. Furthermore, the different embodiments and variants are not mutually exclusive and can be combined with each other. Unless otherwise indicated, the terms “substantially”, “approximately”, “of the order of” mean to within 10%, and preferably within 5%. Furthermore, the terms "between ... and ..." and equivalents mean that the terminals are included, unless otherwise stated.

The invention relates to the characterization of analyte A present gaseous sample to be analyzed. The characterization is carried out by means of an analysis system called 'electronic nose', which comprises: a measuring device; a fluid supply device fluidics; a sensor for the concentration of the parasitic chemical species P; and a processing unit.

As detailed below, the measuring device comprises a functionalized measuring surface defining a plurality of sensitive sites, each sensitive site comprising receptors adapted to interact by adsorption/desorption with the analyte, and here with at least one species chemical P, different from the analyte, and therefore qualified as parasitic insofar as it induces a measurement noise on the signature of the analyte A.

By way of illustration, the electronic nose uses optical measurement technology by silicon interferometric technology (MZI) or by surface plasmon resonance (SFR). The measuring device then comprises an optical source, and at least one optical detector which can be an image sensor or a matrix of photodetectors. The intensity of the detected measurement signal depends on the value of the local refractive index of the sensitive site considered, which is representative of the interactions between the analyte A to be characterized (and the parasitic chemical species or species P) and the receptors.

As a variant, other measurement technologies can be implemented, such as measurement by electromagnetic resonators of MEMS or NEMS type (for example, described in document EP3184485). More broadly, the measuring device can be of the resistive, piezoelectric, mechanical or acoustic type. In the case of techniques for measuring the resonant frequency of a NEMS or MEMS microresonator, the measurement signal can be an electrical signal representative of the vibration of a microbeam or equivalent.

In general, by characterization is meant obtaining information representative of the interactions of the analyte A contained in the gaseous sample with the receptors of the sensitive sites of the electronic nose. The interactions in question here are events of adsorption and/or desorption of analyte A with the receptors. This information thus forms an interaction pattern, in other words a “signature” of the analyte, this pattern being able to be represented for example in the form of a histogram or a radar diagram. In other words, in the case where the electronic nose comprises M distinct sensitive sites, the signature of the analyte A is a vector of dimension M formed by the scalar representative information, these coming from the useful signal Su _m (t) associated at the sensitive site considered.

The analyte A is at least one chemical species intended to be characterized by the electronic nose, and is present in a gaseous sample. It can be, by way of illustration, bacteria, viruses, proteins, lipids, volatile organic molecules, inorganic compounds, among others. In addition, receptors (ligands) are elements attached to sensitive sites and which exhibit a capacity for interaction with the analyte A, although the chemical and/or physical affinities, denoted I< _A , between the analyte A and the receptors are not initially known. The receptors of the different sensitive sites have different physico-chemical properties, which impact their ability to interact with the analyte A, and thus define the different sensitive sites. They may be, by way of examples, amino acids, peptides, nucleotides, polypeptides, proteins, organic polymers, among others. Analyte A can be a single chemical species, or a set of different chemical species whose relative proportion remains constant (for example to within 5% or 10%) from one gaseous sample to another.

In the context of the invention, the gaseous sample which contains the analyte A also contains at least one so-called parasitic chemical species P because it is distinct from the analyte A to be characterized and which can interact with the receptors. Its interaction by adsorption/desorption with the receptors impacts the measurement signal which can then comprise a useful part associated with the analyte A, and a non-useful part associated with the chemical species P in question and forming a measurement noise. These can be water molecules when the relative humidity of the gaseous sample is not zero, an alcohol (ethanol, butanol, etc.), hydrogen sulphide, among others. However, it appears that the concentration variation of the parasitic species P between the initial phase Phi and the characterization phase Ph2 forms a measurement noise which can degrade the quality of the characterization of the analyte A. The characterization method according to the invention makes it possible to limit or even eliminate this measurement noise.

Figure 2 is a schematic and partial view of an electronic nose 1 with SPR imaging according to one embodiment. The electronic nose 1 according to the invention may be similar to that described with reference to fig.lA and fig.lB, and differs essentially in that it comprises a sensor 9 for the concentration of the chemical species parasite P, located for example in the measuring chamber. In the case where there is a variation in concentration of several parasitic species Pi, P2, etc. between the fluidic injection phases Phi and Ph2, the electronic nose may comprise the same sensor 9, or a plurality of sensors 9, adapted to measure the concentration of each parasitic species P. In the rest of the description, for the sake of clarity, it is considered that there is only one parasitic chemical species P.

The electronic nose 1 is based, in this example, on SPR technology and has in this example the characteristics of the so-called Kretschmann configuration, known to those skilled in the art, without the invention however being limited to this configuration. However, as indicated previously, other measurement techniques can be used, such as measurements the resonance frequency of a functionalized MEMS or NEMS type microresonator. As mentioned above, the electronic nose 1 can also be based on an optical measurement by Mach-Zehnder type interferometry, for example in photonic technology on silicon, as described in the patent application FR2011842 filed on 11/18/2020.

The electronic nose 1 comprises M sensitive sites 6 _m , with m ranging from 1 to M, distinct from each other and located in a measurement chamber 4 intended to receive the gaseous sample to be analyzed, these sensitive sites 6 _m being formed each of the receptors capable of interacting with the analyte A to be characterized (cf. fig.lB) and here with the parasite species P. The sensitive sites 6 _m are distinct from each other, in the sense that they comprise different receptors, in terms of chemical and/or physical affinity with respect to the analyte A to be characterized, and are therefore intended to provide information on the interaction that differs from one sensitive site 6 _m to another. The 6 _m sensitive sites are distinct zones of a surface measuring 5, and can be joined or spaced from each other. The electronic nose 1 can also comprise several identical sensitive sites, for the purpose, for example, of detecting a possible measurement drift and/or of allowing the identification of a defective sensitive site.

The electronic nose comprises a measuring device 3, here of the SPR imaging type, making it possible to quantify the interactions of the chemical species with the receptors, for each sensitive site 6 _m , here by measuring in real time the intensity of a optical measurement signal coming from the sensitive site 6 _m considered, this optical signal here being a reflected part of a primary optical signal emitted by a light source 7. The intensity of the optical measurement signal detected by the optical sensor 8 is directly correlated in particular to the adsorption/desorption interactions of the chemical species with the receptors. In the case of techniques for measuring the resonant frequency of a NEMS or MEMS microresonator, the measurement signal can be an electrical signal representative of the vibration of a microbeam or equivalent.

In the context of a measurement by SPR imaging, the measurement device 3 is adapted to acquire in real time the optical measurement signal S _m (t) originating from all of the sensitive sites 6 _m . Thus, the optical measurement signals S _m (t) coming from the sensitive sites 6 _m in response to the primary optical signal are detected together and in real time, in the form of an image acquired by the same optical sensor 8.

Thus, the optical measuring device 3 comprises a light source 7 adapted to transmit a primary optical signal in the direction of the sensitive sites 6 _m , and to generate surface plasmons at the level of the measuring support 5. The light source 7 may be formed of a light-emitting diode, the emission spectrum of which has an emission peak centered on a central wavelength Â _c . Different optical elements (lenses, polariser, etc.) can be arranged between the light source 7 and the measurement support 5.

The optical measuring device 3 further comprises an optical sensor 8, and here an image sensor, that is to say a matrix optical sensor adapted to collect or detect an image of the optical signal coming from the sensitive sites in response to the primary optical signal. The image sensor 8 is a matrix photodetector, for example a CMOS or CCD sensor. It therefore comprises a matrix of pixels whose spatial resolution is such that, preferably, several pixels acquire the optical measurement signal coming from the same sensitive site 6 _m .

The processing unit (not shown) allows the implementation of the processing operations described below within the framework of the characterization method. It may comprise at least one microprocessor and at least one memory. It is connected to the optical measuring device 3, and more precisely to the image sensor 8. It comprises a programmable processor capable of executing instructions recorded on an information recording medium. It also comprises at least one memory containing the instructions necessary for implementing the characterization method. The memory is also suitable for storing the information calculated at each measurement instant.

As described below, the processing unit is in particular suitable for storing and processing a plurality of so-called elementary images acquired at a given sampling frequency f _e , over a measurement period Dΐ, in order to determine a measurement signal S _m (ti), at the current instant t i , associated with the sensitive site 6 _m . Preferably, the measurement signal S _m (ti) corresponds, at a measurement time t, to the average of the intensity of the optical signal reflected and detected by the image sensor 8 on the pixels associated with the sensitive site 6 _m . The average of the optical intensity detected on the pixels can be carried out for one or more images of the sensitive site 6 _m , as described in detail later.

The fluid supply device 2 is adapted to supply the measurement chamber 4 with a carrier gas alone (ie without the analyte A) during the initial phase Phi, and with a gaseous sample formed from the carrier gas and the analytes during the Ph2 characterization phase. The gaseous sample differs from the carrier gas essentially in that it comprises the analyte A to be characterized: the parasitic species P is present in the gaseous sample, and may or may not also be present in the carrier gas. One or more additional gases may be present, but are odorless in the sense that they substantially induce no response from the electronic nose 1. An example of additional gas present in the second gas sample may be vapor phase diluent. The analyte A can thus be stored in a liquid diluent contained in a reservoir 10. The phase diluent vapor and analyte A are added to the carrier gas (eg moist air) to form the sample gas. Be that as it may, the parasitic species P has a concentration C _P J in the carrier gas which may be non-zero or even zero, and a concentration cp _,f non-zero and different from cp _j in the gaseous sample. There is therefore a non-zero difference Ac _p = cp _,f - C _P J.

In the example of fig.2, the fluid supply device 2 may comprise an inlet 11 of carrier gas, and a reservoir 10 of analyte A. Here, the reservoir 10 contains a diluent in which is located the analyte A. It comprises several fluidic lines which connect the inlet 11 of the carrier gas and the reservoir 10 on the one hand, to the inlet of the measurement chamber 4 on the other hand, and comprises valves and possibly mass flow regulators. It thus makes it possible to supply the measuring chamber 4 with the carrier gas (for example humid air with a concentration cp _j of water) during the initial phase Phi and the purge phase Ph3, and with the gaseous sample (for humid air at cp _,f in water, analyte A, and diluent in the vapor phase) during the Ph2 characterization phase. It can be adapted to ensure that the concentration C _A of the analyte A in the measurement chamber remains constant over time. Furthermore, the electronic nose further comprises a sensor 9 of the concentration cp of the parasitic species P, for example here a relative humidity sensor. The cp concentration sensor 9 can be arranged in the measurement chamber, or upstream or downstream of the latter. It is connected to the processing unit, which is suitable for calculating the variation in concentration Ac _p =c _P,f −c _{P i} (or relative concentration) between the initial phase Phi and the characterization phase Ph2.

However, it appears that the concentration variation c _P of the parasite species P present in the measurement chamber during the fluid injection step (phases Phi and Ph2) forms a measurement noise which degrades the characterization quality. This measurement noise is a noise linked to a non-zero difference in concentration cp within the measurement chamber between the initial phase Phi and the characterization phase Ph2. It is a measurement noise insofar as it results from a temporal variation of a parameter which characterizes the environment inside the measurement chamber and should normally remain stationary over time.

This problem of measurement noise linked to Acp is particularly important when the characterization method is carried out from useful signals Su _m (t), that is to say that it comprises a step of subtracting the reference value S _m ,i (baseline) to the corresponding measurement signal S _m (t). In fact, the purpose of this step consists in ruling out from the characterization of the analyte A the effect associated with their environment and in particular the effect of the carrier gas. However, it appears that this reference value S _mi is representative of the carrier gas during the initial phase Phi, but is no longer necessarily representative of the carrier gas during the characterization phase Ph2 since the physical properties of this carrier gas in the measurement chamber may have changed (variation in the concentration cp of the parasitic species P).

FIG. 3A illustrates three interaction patterns, or signatures, reflecting the characterization of different gaseous samples, this characterization being carried out by a characterization method according to an example of the prior art. These signatures M1, M2 and M3 are here representations in the form of a radar diagram of the (stationary) equilibrium values Su _m,f determined from the sensorgrams Su _m (t) in the steady state of equilibrium Ph2 .2 (see fig.lC). They make it possible to highlight the effect of the variation in concentration cp between the phases of fluidic injection Phi and Ph2, here a relative concentration Acp, on the characterization of the analyte A. To obtain these signatures M1, M2, M3 , the carrier gas is identical for the three tests and corresponds to humid air having an initial relative humidity cp _j of approximately 12%.

A first signature M1 corresponds to a gaseous sample formed from humid air with a relative humidity cp _,f equal to approximately 50% and whose analyte A is butanol molecules. The implementation of the characterization process thus presents a relatively large variation in the relative humidity in the measurement chamber, which here goes from cp _j equal to approximately 12% during the initial phase Phi, to cp _,f equal to 50 % approximately during the characterization phase. Also, the reference value S _mi is determined for the carrier gas (humid air at c _{P i} of 12%) and the equilibrium value is determined for the gaseous sample (humid air at c _P,f of 50% with the analyte A) by subtracting this reference value S _mi . This relative concentration Ac _P thus forms a measurement noise whose effect it is important to limit so that the signature M1 is effectively representative only of the butanol molecules.

A second signature M2 corresponds to a gaseous sample formed from humid air with a relative humidity cp _,f substantially equal to cp _j (ie 12%), and whose analyte A is also butanol molecules. The implementation of the characterization method makes it possible, by subtracting the reference value S _m,i associated with the carrier gas (humid air at cp _j ), and insofar as the variation in relative humidity Acp is zero, to rule out the effect of the gaseous environment to thus characterize the interactions of analyte A with the receptors alone. Also, the signature M2 is representative of analyte A alone since there is no measurement noise associated with a variation in relative humidity Acp. Note that the signature M1 is not superimposed on the signature M2, clearly reflecting the presence of the measurement noise associated with Ac _P in the case of M1. It is therefore important to be able to correct the signature M1 to tend towards the signature M2, which is the only representative of the analyte, even though there is a difference in relative humidity Acp in the measurement chamber between the initial phase Phi and the characterization phase Ph2.

The third signature M3 corresponds to a gaseous sample formed solely of humid air with a relative humidity cp _,f equal to approximately 50%. Here, we measure the impact alone of the variation of the relative Ac _P concentration on the characterization of humid air by the electronic nose, in the absence of analyte. It appears that the increase in the relative concentration Ac _P between the initial phase Phi and the characterization phase Ph2 results in an increase in the variation in reflectivity A%R _m of the sensitive sites 6 _m . Note that the Ml signature (humid air with non-zero Acp and the presence of analyte A) is located between the M2 signature (humid air with zero Acp and the presence of analyte A) and the M3 signature (humid air with Acp non-zero and absence of the analyte A), clearly showing the effect of the measurement noise associated with the non-zero relative concentration Acp on the signature of the analyte A. It is therefore important to be able to limit or even rule out this measurement noise to improve the quality of analyte A characterization.

FIG. 3B illustrates examples of the temporal evolution of the concentration cp of the parasitic species P present in the measurement chamber during the initial phase Phi and the characterization phase Ph2, of the measurement signal S _m ( t) for the sensitive site 6 _m of rank m, and finally of the useful signal Su _m (t).

The left part of the graphs corresponds to the situation in which the concentration cp(t) remains constant between the two fluidic injection phases Phi and Ph2. During the initial phase Phi, only the carrier gas is present in the measurement chamber: the parasitic species P has a concentration of CPJ value, which results in a measurement signal S _m (t) having a reference value _Sm ,i. During phase Ph2, the gaseous sample to be analyzed is introduced into the measurement chamber. Due to the presence of the analyte A and since the concentration c _p (t) remains constant, the measurement signal S _m (t) tends towards a stationary value S _m ,f greater than S _mi . To obtain the useful signal Su _m (t), one subtracts from the measurement signal S _m (t) its initial value S _mi , so that the signature is the vector Su formed of the M values Su _m,f obtained.

The right part of the graphs corresponds to the situation in which the concentration cp(t) varies between the two fluidic injection phases Phi and Ph2. Thus, during the initial phase Phi, only the carrier gas is present in the measurement chamber: the parasitic species P has a concentration of CPJ value, which results in a measurement signal S _m (t) having a value of reference S _m ,i. However, during the Ph2 phase, the introduction of the gaseous sample leads to this that the concentration cp(t) of the parasitic species P varies, here increases up to a value cp _,f , higher than Acp with respect to the reference value CPJ.

This variation in the concentration of the parasite species P results in a variation in the measurement signal S _m (t) during this phase Ph2, which here tends to a value S _m,f greater than AS _m with respect to the case where there is no variation in the concentration c _P (dotted curve). It then follows that the useful signal Su _m (t) presents a measurement noise of a value AS _m which depends on the deviation Ac _P . It is then a question, in order to characterize the analyte A, of evaluating the measurement noise AS _m and of subtracting it from the measurement signal S _m (t) with the reference value S _m,i .

To subtract the measurement noise AS _m (Acp) from the measurement signal S _m (t), one approach may consist in carrying out, prior to the characterization step, a calibration step. Patent application FR1913555 filed on November 29, 2019 describes a characterization method comprising such a calibration step. The preliminary calibration amounts to determining a correction function expressing the evolution of the reference value S _mj of the measurement signal S _m (t) associated with the parasitic species P alone (that is to say without the analyte A) as a function of its concentration cp. Thus, during the characterization step, instead of subtracting from the stationary value S _m,f the reference value S _mj determined during the initial phase Phi in which the concentration cp presents the Cp _,j , we subtract the value S _m,i issued from the correction function corresponding to the effective value cp _,f during phase Ph2.

However, it appears that the affinity k _P of the parasitic species P with the receptors can be impacted by the presence of the analyte A. By way of example, when the parasitic species P is molecules of water (water in the vapor phase) and analyte A are hydrophilic or hydrophobic, analyte A adsorbed on the receptors can induce a force of attraction or repulsion of water molecules, modifying the affinity kp accordingly of the parasite species P. The affinity kp _|A is then noted as being the affinity of the parasite species P in the presence of the analyte A. The calibration step may then not accurately estimate the measurement noise effective during the Ph2 characterization phase insofar as the interactions of the parasite species P with the receptors are influenced by the presence of the analyte A.

Also, the characterization method according to the invention is based on the idea of estimating the measurement noise from the interactions of the parasitic species P with the receptors in the presence of the analyte A, and not, as in the prior calibration approach, in the absence of the analyte A. For this, the characterization method provides for carrying out, initially, acquisitions of the stationary values Su _(n,m)f of the signal useful Su _(n,m) (t) for N different gas samples, which contain the same chemical species, namely the same analyte A and the same species or species parasites P, but for which the relative concentration Ac _P(n) is different from one acquisition n to another n+1. We then obtain a matrix of first signatures SUA,P (ie uncorrected signatures) representative of the interactions of the analyte A and of the parasite species P with the receptors, and we obtain a vector of relative concentration Ac _P of the chemical species P during the N acquisitions. In a second step, an optimization problem is solved, from the matrix of first signatures Su _A,p and the vector of relative concentration Ac _P , making it possible to obtain a matrix of corrected signatures Sucn _A , which are representative of the only interactions of analyte A with the receptors during the N acquisitions. These corrected signatures then provide a characterization of analyte A for each of the gaseous samples.

In the examples of characterization method which follow, this optimization problem can be solved in two stages: this is then the first embodiment (flowcharts of FIG. 4 and FIG. 5). As a variant, it can be solved in one step: this is the second embodiment (flowchart of FIG. 6). Other methods of solving this optimization problem are of course possible.

[0066] FIG. 4 is a flowchart of a method for characterizing analyte A according to a first embodiment, which makes it possible to improve the quality of the characterization of analyte A by reducing or even eliminating the noise of measurement linked to the non-zero relative Ac _P concentration between the Phi and Ph2 injection phases. The analyte A is contained in gaseous samples, the latter also comprising at least one parasitic chemical species P. In this example, the parasitic species P is water molecules, but it may be one or more other parasitic species, such as ethanol and hydrogen sulphide, among others.

As indicated previously, during a first phase 100, N first signatures (uncorrected signatures) of the gaseous samples are acquired, which are therefore representative of the interactions of the analyte A and of the parasite species P with the receptors . Then, during a second phase 200, an optimization problem is solved so as to estimate the contribution of the parasite species P in the first signatures, in order to then obtain the corrected signatures then characterizing the only analyte A.

Phase 100: Acquisition of N first signatures, with N>1, of the gaseous samples containing the analyte A and the parasitic species P.

The following steps 110 to 140 are carried out N times, with N>1, each time for a different gaseous sample, these N gaseous samples containing the analyte A and the parasite species P, and differ from each other at least by the concentration c _P of the parasitic species P during the phase Ph2 injection. Each acquisition phase is denoted by an indicator n, a nonzero integer, ranging from 1 to N. Note that the larger N, the better the quality of the estimation of the water signature and therefore the correction of the first signatures.

During a first step 110, a carrier gas G _(n) is injected into the measurement chamber: this is the fluidic injection phase Phi mentioned above. The carrier gas G _(n) does not contain the analyte A. It may or may not contain the parasitic species P: the concentration c _P(n)i may therefore be non-zero or zero. It may also include other chemical species, but which do not have adsorption/desorption type interactions with the receptors of the 6 _m sensitive sites: these are therefore not so-called parasitic species.

The gaseous sample E _(n) is then injected into the measurement chamber: this is the fluidic injection phase Ph2 mentioned above. The gaseous sample E _(n) of rank n comprises the analyte A in concentration C _A(n) and the parasite species P in concentration cp _(n)f . Here also, the gaseous sample may comprise chemical species which do not interact with the receptors. Also, only the analyte A and the parasitic species P have interactions with the receptors and induce a variation of the measurement signal S _(n,m) (t).

During step 120, in parallel with step 110, the measurement signals S _(n,m) (t) associated with the sensitive sites 6 _m are determined, with m ranging from 1 to M, with M >1. These measurement signals S _(n,m) (t) are therefore representative, during the fluidic injection phase Phi (carrier gas G _(n) alone), of the interactions of the parasitic species P, then, during the fluidic injection phase Ph2 (gaseous sample E _(n) ), of the interactions of the analyte A and the parasite species P.

During step 130, the reference value S _(n,m)i (baseline) associated with the carrier gas during the injection phase Phi is determined, and it is subtracted from the stationary value S _{(n ,m)f} of the measurement signal S _(n,m) (t) determined during the injection phase Ph2. A signature Su _(n,m)f = S(n,m)f - S(n,m)i is thus obtained for each sensitive site m, associated with the gaseous sample E _(h > for rank acquisition not.

During step 140, the value cp _(n)i of the concentration of the species P during the injection phase Phi is measured, then the value cp _(n > _f during the phase of injection Ph2, and the difference Acp _(n) = cp _{(n) f} - cp _{(n) i} of concentration (or relative concentration) is determined.

The previous steps are repeated N times, so as to obtain N first signatures Su _(n,m)f for N ranging from 1 to N, and m ranging from 1 to M. From one iteration to another, therefore between n and n+1, the corresponding gas samples E _(n) and E _(n+i) differ between them essentially by the difference Acp _(n) ¹ Acp _(n+i) : we therefore have N differences Ac _P of different values from each other. Note that the concentration c _A of analyte A may or may not vary from one gas sample to another. During step 150, a matrix of first signatures SUA,P representative of the interactions of the analyte A and of the parasitic species P with the receptors of the sensitive sites is formed. This matrix is of dimension NxM and is formed of the values Su _(n,m)f . In other words, each row of rank n represents the signature associated with the gaseous sample E _(n) , that is to say the M useful measurement values Su _(n,m)f of the sensitive sites 6 _m . A relative concentration vector p formed of the N determined values of the concentration difference Ac _P(n) is also formed. This vector is here a vector Nxl dimension column.

Phase 200: Resolution of an optimization problem so as to obtain N corrected signatures, characterizing the analyte A present in the N gaseous samples.

[0078] In general, we are looking for an estimated solution whose product c _A kJ forms a matrix of corrected signatures Suc _A characterizing the analyte A present in the N gaseous samples, the estimated solution minimizing the objective function and maximizing the objective-function. The variables of this problem optimization are c c _A is a concentration vector of analyte A, of dimension N×1, formed of the N concentration values c _A (n) of the gaseous samples; k _A is an affinity vector of the analyte A, of dimension M, formed of the M values of an affinity of interaction of the analyte A with the receptors of the sensitive sites; k _P|A is a vector of affinity of the parasitic chemical species P, of dimension M, formed of the M values of an affinity of interaction of the parasitic chemical species P with the receptors of the sensitive sites in the presence of l analyte A.

During a step 210, the following optimization problem is solved:

In other words, we determine the solutions, local here, minimizing the cost function f such that: This amounts to minimizing the quadratic error, here in the sense of Frobenius norm, between the matrix of first signatures Su _A,p and the term As indicated previously, the terms , are the variables to be optimized, and the terms Su _A,p and Acp are data having been determined during step 150. The function f (and the function g) is called here “cost-function”, but it can be called “cost”, “cost-function”. objective”, “objective”, or even “criterion”. Note that minimizing a cost function f is equivalent to maximizing the function -f. Various known techniques for solving optimization problems can be implemented, for example a gradient descent on all the parameters or a gradient descent by block where each parameter is optimized by considering the other parameters as fixed. . In both cases, it is necessary to start from a random selection of the initial conditions because of the non-convexity of the objective-function which makes it difficult to find the optimum. We then obtain a plurality of local solutions, denoted

In other words, we obtain Q local solutions, each formed of the estimated vectors

[0082] During step 220, the following optimization problem is solved:

This amounts here to determining the solution, among the local solutions previously determined, maximizing the cost function g such that

For this, during step 221, a matrix of corrected signatures Suc _A ^(q) is determined for each of the Q local solutions: Thus, by subtracting the estimated contribution A ^ of the parasite species P in the matrix of first signatures

Su _AP , only the useful contribution is kept, i.e. that linked to the analyte A. This is then to determine which solution, among the Q determined local solutions, maximizes the useful contribution (and therefore the cost function g).

Each of the Q corrected signature matrices Suc _A ^(q) is also normalized to thus obtain Q corrected and normalized signature matrices Sucn _A ^(q) . Thus, we calculate, for each of the corrected signatures of rank n

During step 222, it is then determined which solution qf, among the Q local solutions, has a minimum variance score. To do this, a variance score V is calculated for each of the Q matrices Sucn _A ^(q) , and the one which has a minimum score is identified. The variance score here is the trace of the covariance matrix for each of the local solutions, ie the sum of the variance of each sensor m for each local solution q. More precisely, the variance score can be, up to a multiplicative factor: where is an N-dimensional unit vector, and where is an M-dimensional vector where each m-value is the average of the normalized corrected signatures for the site sensitive 6 _m row m. The solution of rank qf, among the Q local solutions, then corresponds to the vectors whose product is equal to the matrix of corrected signatures

Thus, thanks to the determination method according to this first embodiment, the signatures Sucnj ^ have been determined which effectively and only characterize the analyte A for the N gaseous samples, despite the presence of the parasitic species P, and without resorting to a prior calibration step. The characterization process then provides more precise signatures of analyte A since the measurement noise Ac _P kJ _>|A is estimated in the presence of analyte A and not in the absence of the latter (which is the case as part of the prior calibration).

FIG. 5 is a flowchart of a method for characterizing analyte A according to a variant of the first embodiment. This variant differs from the method of FIG. 4 essentially by the way of optimizing the cost function g.

The method comprises a phase 100 of acquisition of N first signatures. This phase 100 can be identical or similar to that described with reference to FIG. It can therefore include steps 110 to 150, and is therefore not detailed again here.

It then includes a phase 200 of solving an optimization problem consisting of minimizing the cost function f then maximizing the cost function g. The step 210 of minimizing the cost function f can be identical to that described previously and is not explained again here.

The step 230 consists in maximizing the cost function g from the Q local solutions obtained following step 210. This amounts here to determining the solution, among the local solutions determined previously, maximizing the function -cost g such that g

During step 231, the matrix of corrected signatures SUC _A ^(c,) is determined for each of the Q local solutions: This step is identical to step 221. However, unlike step 221, the normalization of the Q matrices of corrected signatures is not carried out

During step 232, it is then determined which solution (of rank denoted qf), among the Q local solutions, has a maximum norm, here within the meaning of the Frobenius norm. For it, we calculate the norm for each of the Q local solutions, and we identify the one that presents a maximum standard. As before, the solution of rank qf, among the Q. local solutions, then corresponds to the vectors whose product c is equal to the matrix of corrected signatures Suc^ ^qf \ [0094] We then normalize the matrix of corrected signatures

. Thus, thanks to the determination method according to this variant embodiment, we determined the signatures which effectively and only characterize the analyte A for the N gaseous samples without having recourse to a prior calibration step. The characterization process then provides more accurate Analyte A signatures.

FIG. 6 is a flowchart of a method for characterizing analyte A according to a second embodiment. This method differs from the methods of fig.4 and fig.5 essentially by the way of optimizing the two cost functions f and g.

It then includes a phase 300 of solving an optimization problem consisting, at the same time, in minimizing the cost function and maximizing the cost-function g = Su _AP — Ac _P k _P|A ). This amounts to minimizing the following function J:

This optimization problem is solved here by an iterative algorithm with a convergence condition. Thus, after an initialization step 311, steps 312 and 313 are carried out iteratively, until a convergence criterion is verified. One thus does not have to carry out a great number of random initializations, insofar as each initialization makes it possible to lead to the total solution (no local solutions). Note here that, for this, the values of the vector Acp are all positive (increase in the concentration c _P(n) between the fluidic injection phase Phi and the fluidic injection phase Ph2). Thus, during step 311, for the iteration indicator i=0, the iterative algorithm is initialized by assigning an initial value to the three variables _A. These initial values can be chosen in any way.

[00100] During step 312, the value at i+1 (here i=1) of the variable is then determined given the values at i (here i=0) of the variables Here we use the fixed point method, insofar as it appears that writing the Jacobian of the function J as a function of the variable k _P | _A equals zero is an equation of the form ( )

Also, we solve, for example iteratively, the equation: way to obtain the value The method amounts to successively applying the function h until convergence.

[00102] During step 313, the values at i+1 (here i=1) of the variables are then determined, taking into account the value ^ which has just been determined. For this, we proceed here to a singular value decomposition (SVD, for Singular Value Decomposition, in English) of the residual matrix R such that R = Su _AP — Ac _P kp| _A. In this example, N is greater than M: N > M. Thus, the residual matrix R can be factorized in the form: R = UåV ^T , where U is a matrix of dimensions NxM whose first column u _m of rank m is proportional to the concentration vector CA of analyte A, where å is a matrix of dimensions MxM whose diagonal elements (o _m )i _<<M are the singular values of the residual matrix R, and where V is a matrix of dimensions MxM whose first column v _m of rank m is proportional to the affinity vector I<A.

[00103] Also, the approximation of rank 1 of the singular value decomposition of the residual matrix R is equal to

[00104] During step 314, it is determined whether a convergence criterion is verified or not. This may involve comparing the variation between i and i+1 of one and/or the other of the variables with a predefined threshold value. When this variation is greater than this threshold, steps 312 and 313 are repeated by incrementing the indicator i by one unit. On the other hand, when this variation is less than or equal to this threshold, we move on to the next step.

[00105] During step 320, it is then considered that the matrix of corrected signatures Suc _A is defined as follows: ^ where if is the value of the indicator i when the criterion of convergence is verified. Thus, the matrix of corrected signatures is defined as being equal to the matrix of first signatures P from which we have subtracted the estimate of the term representative of the measurement noise (impact of the parasitic species). The matrix SUÇ _{A can then be normalized to obtain the matrix Sucn A} _as indicated above at steps 221 and 232. A for the N gaseous samples without resorting to a prior calibration step. The characterization process then provides more accurate Sucn _A signatures of analyte A. Note also that the method according to this second embodiment has the advantage of obtaining the optimal solution in a single step, insofar as the minimization of the objective function f and the maximization of the objective function g are carried out together to obtain the matrix of corrected signatures SUÇA.

[00107] Note that the deviations Acp _(n) determined during the N acquisitions of phase 100, can present a more or less significant amplitude of variation, this amplitude of variation being defined as the difference between the maximum deviation max(Acp _{( n)} ) and the minimum deviation min(Acp _(n) ) among). For a low variation amplitude, for example of the order of 10%, it is advantageous to use the characterization method according to the second embodiment which then gives more precise results.

[00108] FIG. 7 A illustrates three uncorrected signatures Sui, Su ₂ and Su ₃ representative of the analyte A in the presence of a parasitic species P of different concentrations. In these examples, the analyte A is butanol and the parasitic species P is water in the vapor phase (humid air). The Sui signature (dotted line) corresponds to an Acp _,i deviation of 36.5%, the Su ₂ signature (dashes) corresponds to an ACP _,2 deviation of 33.5%, and the Su ₃ signature (solid line) corresponds to an ACP _, 2 deviation of ₃ by -1%.

[00109] FIG. 7B illustrates two corrected signatures Sui and Suc ₂ obtained by correction of the signatures Sui and Su ₂ by means of the second phase 300 of the characterization method according to the second embodiment. The uncorrected Su ₃ signature is reproduced here for comparison (but is not used in the optimization). Note that a low Acp deviation has a low impact on the signature Su due to the subtraction of the reference value S _m,f . It appears that the corrected signatures Suci (dotted lines) and Suc ₂ (dashes) are significantly confused and close to the signature Su ₃ (solid line), thus illustrating the fact that the impact related to the interactions of the parasite species P with the receivers has been fixed.

[00110] Particular embodiments have just been described. Various variations and modifications will occur to those skilled in the art. [00111] Thus, as indicated above, the gaseous samples used during the acquisition phase 100 can comprise several parasitic chemical species: Pi, P2, etc. Also, during step 140, the concentration of these different parasitic species during the injection phases Phi and Ph2, then the deviations AC _PI , AC _P 2 ... are determined. The correction phases 200 and 300 are then similar to those described above, and are based on the optimization of the functions- objective

Claims

1. Method for characterizing an analyte A present in a gaseous sample located in contact with a measurement surface of an electronic nose, the measurement surface comprising M sensitive sites distinct from each other, of rank m ranging from 1 to M, having receptors suitable for interacting by adsorption/desorption with the analyte A and with at least one so-called parasitic chemical species P present in the gaseous sample, the method comprising the following steps: o fluidic injection (110), for bring into contact with the measurement surface:

• during a first phase Phi, a carrier gas which may contain the parasitic species P with a concentration c _P(n)i ; then

• during a second phase Ph2, the gaseous sample containing the analyte A in concentration C _A(n) and the parasitic species P in concentration cp _(n)f , the value cp _(n)f presenting a difference Acp _{(n )} non-zero with the value c _P(n)i ; o determination (120) of a measurement signal S _(n,m) (t), during the fluidic injection step, representative of the interactions of the analyte A and the parasitic species P with the receptors , for each of the M sensitive sites; o determination (130) of first signatures Su _(n,m)f from the measurement signal S _(n,m) (tePh2) associated with the gaseous sample, which having been corrected by a reference value S _{(n, m)i} from the measurement signal S _(n,m) (te Phl) associated with the carrier gas; o determination (140) of a deviation Ac _P(n) of concentration of the parasitic species P between the first phase Phi and the second phase Ph2; o repetition of the previous steps, by incrementing the rank n until obtaining N first signatures representative of the interactions of the analyte A and the parasitic species P with the receptors, with N>1, the gaseous samples being such that the deviations Ac _P(n) are different in pairs; o formation (150) of a matrix of first signatures SUA,P, of dimensions NxM, formed from the N first signatures Su _(n,m)f determined for the M sensitive sites; and a vector of relative concentration Ac _P , of dimension N×1, from the N deviations Acp _(n) determined; o determination (200; 300) of an estimated solution {k _P|A ; _cA ; _kA }; whose product c _A kJ forms a matrix of so-called corrected signatures SUÇA characterizing the analyte A present in the N gaseous samples, the estimated solution:

• minimizing the cost function

• maximizing the cost function are variables defined as follows:

C _A is a concentration vector of analyte A, of dimension N, formed of the N concentration values C _A(P) of the gaseous samples, k _A is an affinity vector of analyte A, of dimension M, formed from the M values of an affinity of interaction of the analyte A with the receptors of the sensitive sites, is an affinity vector of the parasite chemical species P, of dimension M, formed from the M values of an affinity of interaction of the parasitic chemical species P with the receptors of the sensitive sites in the presence of the analyte A.

2. Characterization method according to claim 1, in which the step of determining (300) the estimated solution performs the minimization of the objective function f and the maximization of the objective function g at the same time, the values of the relative concentration vector all being positive.

3. Characterization method according to claim 2, in which the step of determining (300) the estimated solution is performed by an iterative algorithm, with iteration indicator i.

4. Characterization method according to claim 3, in which the step of determining (300) the estimated solution comprises a sub-step of determining the value of the variable k _P|A , given the value of variable c _A and the value of variable k _A , by a fixed point method.

5. Characterization method according to claim 4, in which the step of determining

(300) of the estimated solution comprises a sub-step of determining the value of the variable c _A and the value of the variable k _A , taking into account the value of the variable k _P|A having been determined, by a singular value decomposition of a matrix R =

6. Characterization method according to claim 1, in which the step of determining (200) the estimated solution firstly comprises a sub-step (210) of minimizing the objective function f to obtain a plurality of Q local solutions minimizing the objective function f, with Q>1, followed by a sub-step (220) of maximizing the objective function g.

7. Characterization method according to claim 6, in which the minimization sub-step (210) provides Q local solutions formed each of the estimates of rank q ranging from 1 to Q, of the variables

8. Characterization method according to claim 7, in which the maximization sub-step (220) comprises the determination (221) of Q so-called corrected signature matrices such as:

9. Characterization method according to claim 8, comprising, following the determination (221) of the Q matrices of corrected signatures, a normalization of each matrices of corrected signatures to obtain Q matrices of corrected signatures and normalized

10. Characterization method according to claim 9, comprising, following the determination (221) of the Q matrices of corrected and normalized signatures a determination of a variance score, for each of the Q matrices the variance score being defined as the trace of the covariance matrix for each of the Q matrices the matrix having a minimum score characterizing the analyte A present in the N gaseous samples.

11. Characterization method according to claim 8, comprising, following the determination (221) of the Q matrices of corrected signatures, a determination of a norm of each of the Q matrices followed by an identification of the matrix having the maximum standard.

12. Characterization method according to claim 11, in which the matrix is normalized, thus providing a matrix characterizing the analyte A present in the N gas samples.

13. Characterization method according to any one of claims 1 to 12, in which the electronic nose (1) comprises a device for measuring interactions by adsorption/desorption of the surface plasmon resonance optical type or of the surface plasmon interferometry type. Mach-Zehnder.

14. Characterization method according to any one of claims 1 to 12, in which the electronic nose (1) comprises a device for measuring interactions by adsorption/desorption of the resistive, piezoelectric, mechanical or acoustic type.

15. Characterization method according to any one of claims 1 to 14, in which the electronic nose (1) comprises a fluid supply device (2) adapted to perform the fluid injection step (110), a device (3) adapted to carry out the step of determining the measurement signal (120), a sensor (9) for measuring and determining the relative concentration of the parasitic chemical species, and a processing unit adapted to put implements the step of determining the estimated solution (200; 300).