WO1997028989A1

WO1997028989A1 - System and method for detecting the presence of and identifying a person or object in a given environment

Info

Publication number: WO1997028989A1
Application number: PCT/FR1997/000253
Authority: WO
Inventors: Jean Leteurtre
Original assignee: Jean Leteurtre
Priority date: 1996-02-07
Filing date: 1997-02-07
Publication date: 1997-08-14
Also published as: FR2744548B1; FR2744548A1; AU1798797A

Abstract

A capacitive measurement system including a set of electrodes, multiplexing means for connecting the electrodes to capacitive measurement electronics, and control and processing means. The set of electrodes consists of one or more electrode groups that each comprise a plurality of substantially juxtaposed and contiguous electrodes forming a corresponding number of electrode mats. The control and processing means and the multiplexing means together assign a function to each electrode, such as a measurement, separation or injection function, although the electrodes do not have a predetermined and unchangeable function. The system may further include a subsystem for sensing mechanical stress on the electrode mats. Said system may be used for the detection and identification of vehicle passengers, for surveillance in room spaces, for assisting the loading of containers, and in chairs and beds for the disabled.

Description

"System and method for detecting the presence of and identifying an individual or a _j ob and in a given environment"

The present invention relates to a system for detecting the presence and identification of a person or an object in a given environment.

This system uses the capacitive means; we will also call it "Presence + Identification Capacitance-Based Detector" (PICD).

We will see raw this system has an original design and uses the capacitive means in a new way.

The sensor of this PICD is a thin multi-layer circuit, generally flexible and which can be placed very discreetly on all or part of the walls of the room or passenger compartment to be protected or under the covering fabric of the seat or bed to be fitted , ...

Several applications are envisaged. 1) room monitored discreetly, invisible and undetectable: is there an intruder in this room, 2, 3 people? if yes, where are they in this room

2) container, the walls of which can be covered with this PICD to help or even automate the loading of the container, the position of the packages being measured by the PICD.

3) motor vehicle seat: is such a seat occupied or not? and if so, by what type of passenger? more or less corpulent adult, children seated or not on a booster, baby installed in his specific seat placed on the vehicle seat facing or back to the road? ... If there is a baby seat, is there Is there a baby in this seat?

If the passenger is an adult or a child, what position did this passenger take? 4) seat or bed for disabled person: the PICD placed (undetectably neither in sight nor on palpation) under the seat of the armchair allows to control the movement of the seat of the armchair (combination of a vertical movement and rotation around the horizontal axis normal to the axis of the chair) for an automatic approach to this seat when the disabled person approaches his chair and stands upright, back to the chair; the disabled person places his hands on the arms of the chair which are integral with the seat and are therefore placed in the immediate vicinity of the disabled person; the disabled person thus ensures his balance and can manually or vocally control the slow and controlled return of the seat to the normal position so that the weight transfer from the disabled person takes place with the desired progressiveness from his legs to his buttocks. And vice versa, to assist the disabled person who wishes to get up from his chair.

5) The PICD can still control the modification of the shapes of a seat or a bed to avoid fatigue resulting from a position kept for too long. These shape modifications can be quasi-static, simply to modify a position, or dynamic to automatically perform a massage in which all the parameters, movements, intensities of these movements and frequencies of these movements can be adjusted according to preset programs or not.

The PICD is described in the description; the implementation of the PICD in the applications listed above is also described with some specific achievements taken as examples illustrating the possibilities of this system. The main use _and ANT application to the detection and identification of passenger of a motor vehicle, what is the state of the art?

STATE OF THE PRIOR ART Various means have been used to fulfill these functions of detecting the presence and identifying the passenger in a motor vehicle; in particular ultrasound, radar, infrared, electrical means. These electrical means, resistive and capacitive, are interesting because the corresponding sensors are discreet, non-intrusive, often made up of a flexible sheet of circuit easily accommodable and concealable under the covering fabric of the seat, or even of the ceiling or of the door. or the armrest of the vehicle. Among the electrical means used, let us note the solution proposed by the company TRW Vehicle Safety Systems Inc. (5,232,243 of 03-08-93); its presence and passenger position measurement sensor, placed under the seat cushion, consists of a piezoelectric film, the electrical characteristic of which changes state depending on whether it is more or less pressed by the weight of the passenger: an electrical circuit maintains a forced oscillation of the piezoelectric film, the frequency of which is modified and the amplitude attenuated, all the more since the weight of the passenger causes greater contact between the structure of the detector and this vibrating film. Another system, analogous to a network of strain gauges, uses the changes in electrical resistance induced by the stress exerted by the weight of the passenger on the seat. In the two preceding cases, it is not a capacitive means of passenger detection.

However, we know the company's patent

Robert BOSCH (PCT / DE92 / 00263) who uses the electric heating elements of the electrically heated seat as electrodes of a capacitive sensor to determine if the seat is occupied or not. Also known are patents JP P 60-232371, JP P 280300 and JP P 61-68960 from 1985 and 86 or DE 363 5644 from 20-10-86. These same authors patented (JP P 4-5766 of 16-01-92 and DE 43 01 000 Al of 22-07-93) a new capacitive detection system with no longer an electrode plus the bodywork as previously claimed, but with three electrodes, thus creating not one capacity but three, for the simple purpose of compensating for possible environmental variations. In these patents, the impedance of the human body is considered to be a dielectric of constant equal to 80 (page 4, lines 34 and following of DE3635644 of 20-10-86); this approximation is only true at very high frequencies; indeed, at low frequencies, less than or equal to 1 MegaHertz, the human body is essentially a conductor, whose electrical conductivity is similar to that of an aqueous solution of sodium chloride with a concentration close to 70 grams per liter; it is only around a hundred megahertz that the impedance of the human body has resistive and capacitive terms which are of the same order of magnitude; it is necessary to go up to 1 GHz (1 gigaHertz) and more so that the impedance of the human body is essentially capacitive.

The useful frequencies for a PICD system are around a few kiloHertz: then, the bandwidth of the measurements is sufficient, the line effects are low and the cost of producing the capacitive detection electronics is modest.

In addition, according to the present state of the art, no passenger presence detector is capable of surely detecting the presence of a baby placed in its seat, whatever this seat (some of which are made entirely of plastic material , others have a metal frame). However, it is imperative to infallibly detect the presence of a baby and to inhibit the triggering of the airbag if this baby is placed in its rear-facing seat; the presence detector must also inhibit the triggering of the airbag if the seat is unoccupied; on the other hand, it is necessary to validate the triggering of the airbag in all other cases: seat occupied by an adult or by a child facing the road, but it is desirable to modulate the power of the airbag according to the mass of the passenger and can also be of its position at the time of the shock.

The present invention provides a solution to the problem of detecting the presence and identifying, by capacitive means, of any type of passenger in a motor vehicle, in particular babies placed in their specific seat.

We will now describe the requirements that any PICD system must meet, then describe the means according to the invention which satisfies said requirements. First, a robust and reliable capacitive detection implies the existence of an alternative potential difference between the passenger (who is the target) and each of the measuring electrodes. The passenger's potential must therefore be, at least approximately, fixed. Now there can be no question of defining the electric potential of the passenger of a motor vehicle seated in his seat as can be that of a death row inmate laced on the electric chair; the passenger's potential can therefore only be defined by the electrostatic influence.

Secondly, it is clear that an electrode can only provide information on the distance between this electrode and its target placed opposite. It is not an electrode, nor three, nor even ten electrodes which make it possible to deal with the problem posed, that is to say pretending to identify the many possible cases; the different baby seats and their two positions facing or backward facing, empty or occupied by their baby passenger, children passengers (with or without booster) or more or less heavy adults, the various possible positions of these passengers, that's it so many unknowns to be resolved; and this resolution is only possible if we has a sufficient number of non-linearly dependent data.

We can count about sixty cases concerning babies (15 seat models, positions facing or backward facing, empty or occupied seat) and a good forty cases concerning children and adults (various masses and positions). It is therefore a hundred possible cases of the "passenger" whose identification requires as many independent measures, ie useful information because different from each other, that the PICD must provide, to be able to '' identify the signature of any passenger. Does this mean that at least a hundred sensors are required to validly detect all possible passengers? According to the state of the art, the answer is yes; on the additional condition that these numerous sensors are not used simultaneously, but sequentially: indeed, each sensor, the sensor (1) for example, forms with the passenger a capacitor whose capacity, Ci, is variable according to the passenger and its position; the passenger also has capacity C with the bodywork and this capacity C is also variable depending on the passenger, the position of the passenger on his seat and the configuration of this seat, more or less advanced and / or reclined, in the passenger car considered . The passenger potential,, is balanced between zero, which is the body potential, and the alternative potential Vo (f °), (of frequency f ° and amplitude Vo), used to perform the capacitive measurement. If all the sensors were simultaneously measuring, the sum C of hundreds of capacities Ci would be much greater than C: the passenger's potential would then be so close to the alternating potential Vo (f °) that the capacitive detection thus carried out could only be poor. , for the reason mentioned above.

If the hundred or more sensors are used sequentially, thanks to a multiplexer which switches the electrodes of all the other sensors to zero potential, then the sum C of the hundreds of capacities Ci, minus the capacity Cj of the sensor j being measured, the more the capacity C, that is to say all the capacities that the passenger has with surfaces at zero potential, greatly exceed Cj which influences the passenger towards the potential Vo (f °); therefore the potential of the passenger,, takes a value very close to zero; and the capacitive measurement would then be correct; but the cost of such an implementation would be high because of the very high number (a few hundred) of sensors and also because the multiplexer must have a coaxial structure; but this hypothetical passenger detector according to the state of the art (ie frozen, each of its electrodes having a defined function) would not be capable of the variable focusing which will be described later and which is the source of the richness of the PICD measurements according to the invention.

Capacitive detection systems have the disadvantage of being able to be blinded, and rendered inoperative, if the seat is flooded (bottle overturned, rain, pee ...),

- or if the seat is covered with a cushion containing pure coil springs allowing air circulation between seat and passenger,

- or even if the passenger has between the seat and him, a magnetic mat for his psychological comfort. The object of the invention is to propose a system for detecting the presence of a person seated in a passenger compartment, in particular a passenger in a motor vehicle, which can provide real passenger identification information while taking account of external disturbances. any aforementioned.

These disturbances are circumvented by doubling the multiple capacitive detections between the sensor electrodes and the passenger (which constitutes the non-contact measurement subset, entitled "capacitive identification subset 1") by a second subset which no longer measures passenger-sensor distances, but the constraints applied by the passenger on the various parts of the seat; this second subset of the PICD is called "applied stress measurement subset 2". Thanks to these two subsets of measurements, our PICD according to the invention generally offers complete information on the passenger; and if the subset 1 is temporarily blinded therefore inoperative, the subset 2, which is insensitive to the above-mentioned disturbances, always provides the minimum information which allows the decision to authorize or inhibit the firing of the airbag associated with the seat in question.

In the system according to the invention, the two sub-assemblies use the capacitive means, the sub-assembly 2 measuring the stresses applied by:

- the capacitive measurement of the crushing of a rubber film, or

- the measurement of the air pressure in separate air cushion elements for a predefined crushing (measured capacitively) of each of the small elementary cushions of which the assembly forms the seat on the one hand and the backrest on the other go.

The method and system of capacitive measurements according to

1 invention claimed here solve the following three paradoxes.

- the first paradox is to solve the hundred unknowns detailed above, with fewer sensors than the number of unknowns,

- the second paradox is that the invention allows more wealth of information than the solution according to the state of the art as described above, despite its number of electrodes lower than the number of unknowns, it is to say of possible cases to be resolved, - the third paradox lies in the cost of implementation significantly less than that of the solution according to the state of the art as described above. In addition, the method or the system according to the invention is safe insofar as it always provides the minimum information necessary for the decision to trigger or prevent the firing of the airbags, whatever the accidental disturbances described above. .

Finally, the detection systems according to the invention can also be advantageously implemented for security purposes to eliminate or limit the effects of "whiplash" by providing that the stress detection subsystem is arranged to detect a excessive pressure exerted by a passenger on the backrest in the event of a rear impact on the vehicle.

In this version of the detection system according to the invention, the capacitive identification subsystem is arranged to detect the position of the passenger's head relative to the backrest. The detection system according to the invention is then preferably arranged to control the triggering of an air cushion device (airbag) placed at the top of the backrest or in the headrest to retain the passenger's head, in detection of a position too far from the head in relation to the backrest.

DESCRIPTION OF THE INVENTION The method and / or the system of capacitive measurements according to the invention will now be detailed, with the help of FIGS. 1 to 20:

- Figure 1 shows the seat of a motor vehicle equipped with two sensor mats one for the seat, the other for the backrest and their connecting tracks to the electronics integrated in the seat structure, -

- Figure 2 shows the block diagram of the elements constituting the electrode mat; - Figure 3 is the block diagram of this electrode mat, the electronics 40 which is preferably an ASIC grouping all the necessary functions and the connector 52 for connection between 40 and 41; - Figures 4.1, 4.2 and 4.3 show a real example of electrode pads 12 and 14 with their tracks and the connector 52: 4.1 for the seat 41, 4.2 for the backrest 42, 4.3 for the detail of the connector 52 for the tracks and guard tabs separating each track;

- Figure 5 is the diagram of the switch 53 whose role is to allow, under control of the processor 58, to connect each electrode 50 to one of the three potentials 55, 56 or 57; - Figure 6 illustrates, on the example of a hexagonal cross-linked network of 63 electrodes, the richness of the possibilities of transformation of such a multisensor.

- Figures 7 and 8 give examples of electrode configuration; the following codes are adopted: 0 if the electrodes have no assigned function, 1 for those belonging to R, 2 for those belonging to G, 3 for those belonging to E;

- Figure 9 is the count of the number of different measurements from each other, calculated for all the configurations shown in Figure 8 and for three dimensions of the mat 41: v = 5, w = 4; v = 6, w = 6; v = 8, w = 8;

- Figures 10, 11, 12 illustrate all the possible measurements, in a network u = 8, w = 8 for the configurations 12 _3ι4 -4-2 _/ l _{1> 1} -4-4, l _1/1 -5- 4; - Figure 13 shows the preferred arrangement of the two electrode pads for the two PICD subsystems and the rear guard;

FIG. 14 shows an example of a particular measurement of the subsystem 1 carried out at time t and of a measurement carried out by the subsystem 2 for an electrode situated opposite an electrode E of the subsystem 1 at time ( t + dt);

- Figures 15.1, 15.2 and 15.3 repeat and complete the example of Figures 4.1 to 4.3; - Figure 16 shows an element of 12 electrodes, reduced (Figure Iδbis) to three groupings of the electrodes four by four; - Figures 17 show a preferred version of the PICD specially adapted for detecting a rear impact and preventing whiplash;

- Figure 18 is the diagram of the preferred version of 40 with the two capacitive bridges 54 and 64 specifically added to treat the prevention of whiplash.

- Figures 19.1 and 19.2 illustrate examples of airbags instead and in place (function) of the elastic dielectric layer 25;

FIG. 20 shows the functional diagram of the inflation of these cushions with their solenoid valves operating in closed loop under the control of the sensors of the subsystem 2, or that of the processor 58 for a program of static change of position or of dynamic change preprogrammed and called by the passenger via processor 58.

DETAILED DESCRIPTION OF A PARTICULAR EMBODIMENT

PICD As announced above, the PICD according to the invention can fulfill two functions:

the first series of measurements is a large number of capacitive measurements between an electrode mat 50 placed under the covering fabric of the seat 1 and the body of the passenger; these are contactless measurements; the electrodes of this first subsystem and their electrical supply tracks are located in the first conductive layer 12 of a flexible multilayer circuit 10 placed between the foam of the seat 2 and the fabric 3 of the cover of this seat (Figures 1 and 2).

- The second function provided by the PICD is the measurement of the distribution of the loads exerted by the passenger on the seat, seat and back. This measurement of the stresses applied to the seat surface by the passenger is carried out capacitively by a second layer of electrodes 14 aiming through a dielectric layer 25 another conductive layer 15, both placed inside the flexible multilayer circuit 10 and having between them an alternative potential difference for carrying out the capacitive measurement of the distance between these conductors 14 and 15. These layers ₁ 4, 15 and 25 being buried in the thickness of the flexible and sealed circuit 10, no external disturbance such those described above (flooded seat ...) cannot disturb this second series of stress measurements applied by the passage (or a package) placed on the seat, hence the safety of these measurements which supplement those carried out by the first sub -system.

FIG. 3 schematically shows the PICD, the main components of which are the electronic assembly 40 and the two flexible multilayer circuits 41, 42 connected to the electronic assembly 40 by extensions 41bis, 42bis. The two multilayer circuits 41, 42 are respectively assigned to the seat and the backrest, which form two distinct parts of the seat.

The multilayer circuit 10 in FIG. 2 is the functional diagram of the various functions requested, with the layers of measuring electrodes 12 and 14 separated by a guard plane 13, the layer 15 targeted by the electrodes of 14 and even a possible layer 16 , heated for the comfort of the passenger in cold countries and also to dry the seat faster, the fabric 3 of which would have been accidentally flooded.

It will be seen later that the multilayer circuits 41 and 42 fulfill exactly the same functions as the circuit 10 in FIG. 2 but are in reality much simpler. Let us now see the principle of capacitive measurements which has been implemented for this PICD. We will see that the principle of these capacitive measurements is original and that the design of the sensor is also original; it is the least after announcing that this PICD according to the invention resolves several a priori surprising paradoxes.

The first essential characteristic of the invention is that the electrodes 50 of the sensors are adjacent to each other as much as possible (Figure 3); the electrodes form a mat 41 of rectangles or hexagons or of any other quasi-contiguous figures, without empty portion, on all the surfaces of 41 except those reserved for the seams of the covering fabric; there is no need for the electrodes to have the same surface; they can have non-geometric shapes, in particular on the periphery of 41; if the pad of electrodes 41 is intended to cover the passenger seat, it is advisable to give each of the peripheral electrodes 50 the obviously non-geometric shape, which makes it possible to match the shapes of the seat and the backrest.

The second essential characteristic of the invention is that the electrodes of the sensors do not have a defined, immutable function, as is the case according to the prior art, where there is for each sensor a measurement electrode and an electrode for guard surrounding the measuring electrode on all sides, except for the front face which is the measuring side. The novelty characterizing the invention is that each electrode can be switched, under the control of a processor, to one of the three potentials presented by all capacitive measurement electronics using the diagram of the charge amplifier, namely: "the input 55 of the operational charge amplifier (tab -), connected to the measuring electrode, and whose potential is Vm,

the potential 56, which is that of the tab + of the operational charge amplifier and which is connected to the guard, is equal to Vm, in amplitude and phase,

- potential 57, whose potential V differs from Vm by Vo (f °):

Vm - Vi = Vo (f °) (1), which is used to influence the potential of the passenger, or more generally that of any target whose potential cannot be fixed by a driver, as detailed above. A single capacitive measurement electronics (54) generates the three functional potentials 55, 56 and 57, defined above; a network of switches 53, controlled by the process 58, makes it possible to switch each of the three functional potentials 55, 56, 57, to each of the electrodes 50, according to a series of programs recorded in 58.

The most natural location for the electrode mat 41 is the seat; because seat and back are two separate parts of the seat, here we give the reference 41 for the electrode mat of the seat and 42 for the electrode mat of the back.

Other locations of the electrode mat is the passenger compartment ceiling, the dashboard, the interior trim of the door adjacent to the passenger: it is obvious that the passenger can be all the more finely characterized as there are more sensors arranged in the surfaces surrounding this passenger.

If we are only interested in the equipment of the seat, the tracks 51 of the electrodes of 41 and 42 arrive on the terminal block 52 where the electrical junction of the electrodes 50 and the electronic circuit 40 is carried out; the preferred solution for 40 is an ASIC integrated circuit which groups together the essentials of functions 54, 58 and 53 in one housing. The terminal block-connector 52 is also the receptacle for ASIC 40, also described in the patent application. no.9613992.

Figures 4 show an example of the electrode pads that can be used to equip the seat of a Renault Mégane, Figure 4.1 showing the drawing of the seat layer 12; Figure 4.2 for the same layer 12 of the backrest, with the two locations reserved for seams, while the seat has only one seam. The true width of the electrodes is 30 cm for the prototypes shown in Figures 4.1 and 4.2.

In addition to the connection (s) of the tracks 51, the circuit 40 has the 0 and 12 Volt supply terminals and some outputs, those of the authorization signal or inhibition of the firing of the airbag (s): inhibition in the presence of a baby placed back to the road and in the absence of a passenger (empty seat), etc .; an authorization for the headrest airbag in the event of a rear impact; another output is the alarm signal warning the driver of a possible malfunction of the PICD, for example the seat flooded and not yet dried by 16 (see Figure 2).

The arrangement of the PICD sensor electrodes and, on the other hand, the possibility of choosing the potential distributed to each of these electrodes constitute, together because these two features are inseparable from each other, the essence of the invention; the interest of this association will now be explained.

DESCRIPTION OF THE OPERATION OF THE PICD FOR THE SUFFI-

SYSTEM 1 Thanks to the switches 53, the N electrodes (Figures 3 or 4) can be connected indifferently to one of the three terminals 55, 56, 57. The N electrodes 50, distributed in three subsets R, G, or E , are used sequentially; R (or M) mean receiver (or measure); G is taken for guard; E (or I) mean emitter (or charge injection. By convention, in Figures 7 to 12, the electrodes of R are represented by 1; those of G by 2 and those of E by 3. Only a small number n ( n = l, or 2, or 3, ...) of these N electrodes forms the subset R, these n electrodes being successively taken from among the N electrodes; all the p electrodes adjacent to the n electrodes (as first neighbors or seconds or third neighbors of the n electrodes) are switched to 56 and form the subset G; the electrodes of groups R and G are brought to the same potential: Vm, which is the potential of the input of the charge amplifier, - the guard is brought to the same potential Vm. All the other electrodes form the subset E: these are the (Nnp) electrodes brought to the potential Vi and whose role consists of electrostatically influencing the passenger's potential.

The printed circuit 41 is a very special multi-sensor, because it is multifaceted; this multi-sensor not being frozen, its N electrodes can be selected, in rapid sequence, to be placed successively in the R group (measure) always keeping the G group around the R group, like the Monarch's Close Guard which moves in the N squares of his kingdom: when R occupies a peripheral square, he is no longer guarded on all sides by elements of the checkerboard of electrodes; however, it will be seen later that the preferred PICD solution according to the invention nevertheless provides adequate guarding for M, when R occupies one or more peripheral cell (s).

In a motor vehicle, all electrical potentials are related to the potential of the bodywork, taken as zero. Several capacitive measurement solutions are possible; for example and in accordance with relation (1),

- a): Vm = V ₀ (f °) and Vi = 0,

- b): Vm = + V ₀ (f °) / 2 and Vi = -V ₀ (f °) / 2, which is the same potential, phase shifted by 180 °,

-c): Vm = 0 and Vi = V „(f In the case a. the (n + p) measuring and guard electrodes are brought to potential V _o (f ⁰ ); then the bodywork, by definition at zero potential , adds its contribution C to the sum of the capacities Ci of the (Nnp) injection electrodes (which are at potential Vi = zero in this case); the potential of the passenger, v, is defined by equation (2): i = Nnp I = n I = p (C + Σ (Ci)) xva = (Σ (Ci) + Σ (Ci)) x (V _o (f ⁰ ) -v) (2) 1 = 1 1 = 1 1 = 1

1 = N-n-p I = n l = p

By calling C _E = Σ (Ci), C _R = Σ (Ci) and C _G = Σ (Ci) (3) 1 = 1 1 = 1 1 = 1 (2) becomes: (C + C _E ) χv _a = (C _R + C _G ) x (V ₀ (f °) -v _a ); the potential goes from the passenger is: v _a = (V ₀ (f °) x (C _R + C _G ) / (C + C _R + C _G + C _E ) (4). Each new distribution of the N electrodes in three subsets R, G, or E brings new values for the sums of capacities, C _E , C _R , C _G , and therefore defines a new value of the passenger's potential, v _a , in the case a; to 1A measure where C _E is most often significantly greater than C _R + C _G , the potential v _a of the passenger is often closer to zero than V ₀ (f °) in this case a ..

Conversely, in case c, the equilibrium equation of charges becomes: (C + C _R + C _G ) xv _c = (C _E ) x (V ₀ (f °) -v _c ) (5) and then the potential v _c of the passenger, which also fluctuates with each change in the position of the passenger and with each change in the distribution of potentials on the electrodes, is most often closer to V ₀ (f °) than to zero; its expression is: v _c = (V _Q (f °) χC _E / (C + C _R + C _G + C _E ) (6).

Whatever the capacitive measurement solution chosen, one of the three particular cases described above or others, the potential v of the passenger is only a fraction of the alternating potential V _o (f ⁰ ), of which l amplitude is only a few volts: the PICD is completely harmless and poses no risk to the passenger.

On the other hand, whatever the capacitive measurement solution chosen, the capacitive measurement electronics 54 delivers an output voltage Vs, continuous, proportional to the product C _R χv. Hence the expressions derived only for the special cases a and c (the other cases giving similar expressions of Vs) • Vs _a = KV ₀ (f °) x (C, x (C, + C _G ) / (C + C _R + C _G + C _E )) (7)

Vs.≈KV ₀ (f °) x (C _ε; xC _E ) / (C + C _R + C _G + C _E )) (8).

Expressions which clearly show the complex implication of the distribution of the electrodes of the three R, G and E subgroups in the result of each measurement given by the PICD; (C-rC _R + C _G + C _E ) is the overall passenger capacity with the PICD and the passenger compartment of the vehicle.

The method and / or system according to the invention, that is to say this new arrangement of electrodes and this new way of rapidly defining the function of these electrodes, make it possible to make good capacitive measurements because the potential v of passenger is well defined for each measurement; even if the passenger's potential changes with each configuration, that is to say each millisecond.

This multi-sensor, with this way of using it, presents an analogy with an optical instrument with variable focal length: the dimension of the measuring electrode (n = l or 2 or 3, etc.) defines the size and the number of pixels; the extent of G, variable with the number p, is the analog of variable focusing. In addition, the E domain plays the role of illumination, or rather of repetitive flash, since it is typically every milliseconds that a new "photographer" is taken by the PICD, and this with a different "illumination" each time .

The limit is encountered when the potential v of the passenger ceases to be correctly defined, or to continue the optical analogy, when the "lighting" of the scene becomes too weak, even random. Although formulas (7) and (8) show that the result of a measurement by the PICD is a function of the capacities C _R , C _G , C _E , and not of the numbers n, p or N, the various possible combinations of the PICD will be classified according to the number without dimension, nor significance, τ≤ (n + p) / N; τ only gives the ratio of the areas of 41 carried to the two different potentials Vi and Vm, which both electrostatically influence the passenger. Clearly, it is not possible for n + p to exceed N; but it would be possible that n + p = N: then the domain E would no longer exist and the passenger's "lighting" would only be provided by capacity C, except for the preferred solution c; the capacitive measurement would then become impossible, since only the electrodes of E "light" the passenger. All cases of τ = l will be deducted from the number of combinations of electrodes that can be used. In addition to the excellent quality of the capacitive measurements carried out thanks to the good definition of the potential v of the passenger, a new considerable advantage of the PICD is to enormously multiply the total number of combinations of electrodes; Now all these combinations of electrodes are as many different capacitive measurements, because on the one hand the geographical location and the extent of the domain R and on the other hand the extent of the domain G which is peripheral to it define, together, the electric field lines exchanged between R (which changes every millisecond) and the passenger (a hundred possible cases and which can move). The larger the area G, the more electric field lines from R which go to the passenger; but also, the more the R domain increases, the better it is possible to detect the passenger at distances which are all the greater as the two R and G domains are themselves larger.

A few examples to illustrate the above A preferred version of the PICD according to the invention will now be described. So far in the description, no specific capacitive measurement solution has been preferred. In all cases, the switch 53 has the structure shown in FIG. 5, each electrode 50 is connected by its track 51 to its switch 531; the sources of potentials 55 and 56 are connected to switch 532, while the sources of potentials 57 and 56 are connected to switch 533; this arrangement is necessary to avoid a direct view of the potentials 55 and 57 - when the electrode 50 is connected to 55 (position R), 532 is positioned so that 56 is switched on the lug 531-2, as shown in the figure 5; This avoids the leakage capacity (between terminals 531-1 and 531-2) of several picoFarad which would desirably be added to the measurement Alternatively, if 50 is connected to 57 (position E), the switch 532 links 531-1 to 56, again to avoid a direct view of the potentials 55 and 57 in the switch 531.

Three particular capacitive measurement solutions, offering the same ease of implementation, have been presented previously. Solution c is however preferred for all of the following reasons:

- if the potential Vm of the input of the charge amplifier, like that of the guard is zero, that is to say the same potential as that of the bodywork, the latter then plays the role of electrical guard, this which not only simplifies and the realization of the sensor 41, and that of the possible connections or extensions 41bis, and especially that of the switches 53, but also has the additional advantage that the bodywork provides a guard for the electrodes of R when it (s) occupies (nt) one or more peripheral case (s) of 41;

- formulas (7) and (8) detail an additional advantage of c compared to a, taken here as an example of a solution different from c: the measurement ratio is: Vs _c / Vs _a = C _E / (C _R + C _G ) (9) Already, Cj is very often greater than (C _R + C _G ); but above all, for the detection of baby seats, in particular the most delicate case of a seat with a metal frame, the baby-domain E coupling is significantly better; in the optical analogy of the page, the baby is well "lit" in this preferred electronic solution.

It now remains to calculate, for a few particular cases, the gain provided by the concept according to the invention.

Let us carry out these calculations by considering a carpet of rectangular electrodes forming a matrix of v columns and w rows, and several values of v and w: v = 5, w = 4 v = 6, w = 6 v = 8, w = 8 . The number of different capacitive measurement possibilities is indeed considerable; this is now detailed, with Figures 7 and following:

- with n = l, p = 8 if the guard has a row of electrodes (except if the measuring electrode is placed at the edge of the circuit 41), the scanning of this measuring electrode on the set of N electrodes allows filling a matrix vxw of capacitive measurement results, with 1 ranging between 6.3% and 14.1% (the minimum and maximum values of are henceforth calculated for v = w = 8), the index recalling the number of electrodes forming the guard width; a second matrix vxw of results is filled if the guard can have up to two rows of electrodes, with 2 ranging between 14.1% and 39.1% (p is then 24), a third matrix vxw of results if the guard can have up to three rows of electrodes, with 3 between 25% and 76.6% (p is then 48), etc.

- Let us continue the multiplexing of the electrodes by taking n = 2: depending on whether these two electrodes are grouped in rows, or in columns, or in diagonals, we can fill a matrix vx (w-l), a matrix

(vl) xw and two matrices (vl) x (wl) new results of capacitive measurements if the guard has a row of electrodes, still four other matrices: a matrix vx (wl), a matrix (vl) xw and two matrices (vl) x (w- 1), if the guard has two rows and four other similar matrices if the guard has three rows of electrodes, with the respective extreme values of: 1 = 9.4 - 18.8% n 2 = 18.8 - 46.9% and 3 = 31.3 - 75%. - With n = 3 aligned electrodes, there are as many matrices vx (w-2) and matrices (v-2) xw and again twice as many matrices (v-2) χ (w-2) as there has guard widths which are considered; the extreme values of are respectively 1 = 12.5 - 23.4%, 2 = 23.4 - 54.7%, 3 = 37.5 - 75%, 4 = 39.1 - 100% (for the latter value of 4, the potential of the passage is zero and the passenger's capacitive measurement then becomes impossible, as explained above. - The electrodes of R can also form a triangle (n = 3 or 6 ⁾ , a rectangle (n = 4 or 6 or 8, etc.), a parallelogram (n≈4 or 6 or 8, etc.), a centered diamond ( n = 5), etc. - Finally, there is no obligation to consider only guard widths expressed as the number of electrodes equal in all directions, as above: other guard configurations are interesting: the electrodes being rectangular (60mm x 130mm) , four electrode widths are only 240mm while two electrode lengths make a guard of 260mm.

Generalizing, first with the notation

( ⁿ _t ax, ay) -b _x -b _y ) defining the configuration, characterized by the number n of electrodes in position R; the indices a _x and a _y measure the projections of R along X and Y, projections expressed in number of electrodes; b _x and b _y are, in number of electrodes, the projections in X and in Y of G which must surround M, at least on one side in X and on one side in Y. The notation l _{1 (1} - 4-2 means that R has only one electrode and that G has, at least on one side in X, a range b _x = 4 electrodes in the direction X and, at least on one side in Y, b ₂ = 2 electrodes; the notation 3 _{3> 1} -4-2 represents a domain R of 3 electrodes along X, surrounded by a guard of b _x = 4 electrodes at least in a direction of X and of b _y = 2 electrodes in a direction of Y. The notation 4 _2r2 -4-2 represents a domain R of 4 electrodes in rectangle, 2 according to X (a _x = 2) and 2 according to Y

(a _y = 2), whose guard measures, at least on one side in the X direction and on one side in the Y direction, respectively b _x = 4 electrodes and b _y = 2 electrodes.

Each configuration, characterized by the notation (n _{(aX / ay)} --b _x -b _y ), generates the number Np of possible configurations: Np = ((va _x +: ^ - N _x ) ((wa _y + l ) - N _y ) (10)

N _x being the number of configurations impossible because the guard would not have its characteristic number of electrodes around M, at least in an X direction; if (a _x + 2b _x - (v + 1)) is negative or zero, N _x = 0, if (a _x + 2b _x - (v + 1)) takes a positive value, N _x is this value; ditto for the term Ny, which is zero if (a _y + 2b _y - (w + l)) <0, and which is equal to (a _y + 2by - (w + 1)) if (a _y + 2b _y - (w + l))> 0.

To obtain the total number, N, of measures offered by ^' (n (a _x , a _y ) - (b _x -b _y )), we must subtract from N _p the number No of possible cases, but generating the refused situation of τ = l, situation which can only occur if the two conditions: a _x + 2b _x ≥v (11) ej; a _y + 2b _y > w (12), are verified simultaneously; and No is 1, 2 or 4 depending on the values taken by (ax + 2bx) ei of (a _y + 2b _y ): - if a _x + 2b _x <v OJJ. a _y + 2b _y <w, No = 0

- if a _x + 2b _x = v e_ £ a _y + 2b _y = w, a single configuration induces τ = l; No = l,

- if a _x + 2b _x > v e_ £. a _y + 2b _y = w, or if a _x + 2b _x = v ei a _y + 2b _y > w, there is No = 2 conf igurations inducing τ = l,

- if a _x + 2b _x > v e_t a _y + 2b _y > w, No = 4.

Hence the number N of different measures of fert by the configuration (n _{(aX / ay)} - -b _x -b _y ):

N = ((v- a _x + l) - N _x ) (w- a _y + l) - N _y ) - N ₀ (13) The examples which follow illustrate this, in particular Figures 10, 11 and 12. The table of figure 9 recapitulates the number of different measurements for three particular cases chosen arbitrarily by way of examples, namely (v = 5, w = 4), (v = 6, w = 6), (v = 8, w = 8) and for all the configurations shown in FIG. 8. These are respectively 247, 790 and 1730 different measurements counted for the three electrode pads (v = 5, w = 4), (v = 6, w = 6) and (v = 8, w = 8). Even if a certain proportion of these capacitive measurements does not provide significantly new information, Figure 9 clearly shows that the PICD, even the PICD (v = 5, w = 4), offers a number of measures much greater than the number unknown, estimated at a hundred. Detection, identification of the passenger and his position are therefore possible. The problem would rather be now the abundance of these possible measures and, in order not to be cluttered with too much data, to choose judiciously: - the number N of electrodes,

- the number of combinations of electrodes. The manufacturing cost increases, but relatively little, with the number of electrodes. A number N, a little higher than the twenties which seems to be a necessary minimum, offers not only the advantage of a very large glut of possible measures, but also that of being able to benefit from all the richness of the adaptive capacimetry described above. . For example, the PICD prototype 8x8 allowed the testing of around thirty combinations of electrodes; a measurement campaign among others have been conducted with the four combinations the _lfl -4-2, -4-2 _3ιl 3, 3 _{1; 3} -3-l, 4 _2, ₂ -4-2 , which provided for each "passenger" respectively 64, 48, 48, 49 measurements, that is to say 209 different measurements, ie more than necessary to identify the passenger. The detection of the presence of an adult and the identifications of its size (therefore its mass) and of its position are a simple problem since the capacity measured by an electrode on which the passenger is seated is significantly higher than that measured by an electrode more distant from the passenger: a short program of multiplexing the electrodes is sufficient to surely detect the presence of such a passenger and even estimate its size and position (from the contour of the electrodes indicating high capacities); the identification of the passenger kneeling and facing back (passenger giving the bottle, ...) is also easy. Babies are much more difficult to detect, because their seats keep them away from the electrodes, and also because some seats are electrically insulating while others have a metal frame; It is for these more difficult cases that the PICD demonstrates all its investigative power at several distances, close or distant, according to the shapes and dimensions of the R and G domains. In summary, the best use of the new concept and / or of the system according to the invention is a PICD not too minimalist and whose processor 58 is provided with software allowing a first exploration, to quickly identify the presence or absence of a passenger, then this point being acquired, the continuation of the measurement for the only more delicate cases, ie the identification of the position of a baby , back or facing the road, until it is unambiguously identified.

It remains to extract information from the hundreds of measurements programmed and carried out in the preceding instant. These measurements are presented for example in the form of a global matrix, sum of several sub-matrices each corresponding to the multiplexing of a defined configuration of electrodes.

The digital processing of data is the comparison of all or part of this global matrix of measurements with the corresponding elements of the hundreds of matrices, stored, relating to the hundred typical typical cases, previously identified. With 2 to 300 measurements to process and therefore 100 times 2 to 300 values corresponding to the same 2 to 300 measurements stored for the 100 reference cases, a least squares method requires 2 to 30,000 subtractions, 2 to 30,000 squared elevations, 2 to 30,000 additions and a few hundred more additions and subtractions to compare the current measurement with the 100 reference cases; this is entirely feasible, in a short time, by the microprocessor specialized for the application. A neural network method is well suited to more effectively provide the identification of the measurement carried out the previous second among the hundreds of recorded types.

Data processing can be much more simplistic and be only the sum of the output voltages of the single capacitive measurement bridge contained in the ASIC 40, this for some electrode configurations (n _(aXιay) - b _x -b _y ).

The above shows the power of the concept according to the invention for the position detection and measurement subsystem 1 carried out by the electrode mat of the 1st layer (12) of the flexible multisensor 10 or 41

(Figures 2 or 3).

Let us now consider the subsystem 2 for measuring the stresses applied by the passenger on the surface of the seat. The subsystem 2 is much simpler: it is the simple capacitive measurement of the crushing (induced by the weight of the passenger) of the elastic and dielectric layer 25 placed between the electrodes 14 and 15 of FIG. 2. Alternatively, it is the measurement of the pressure which must be given to the fluid dielectric 25 contained in n independent and adjacent bags, the assembly of which forms an airbag, such as that shown diagrammatically in FIGS. 19.1, 19.2 and 20.

In the example described in Figure 19.1, the bags 30 are thin, each bag receiving a controlled pressure so that the capacity, given the distance, between 21 and 23 is constant, for example of the order of

5 to 10 mm.

In the example described in Figure 19.2, the bags 30 can have a thickness of the order of 5 to 10 cm; the seat is no longer necessarily fitted with flexible foam; the bags 30 form an airbag; they almost all receive the same pressure, at least by areas: backrest, seat under the pelvis and seat under the legs. The passenger thus has excellent comfort (such as an anti-decubitus seat) and also experiences little or no vibration from the vehicle.

Before developing the description of this alternative, let us finish with regard to the capacitive measurement of subsystem 2. We now describe a preferred solution of the PICD, with the electronic assembly

40 associated with the preferred multisensor 41. We will consider the multisensor 41 as a whole, whereas it is actually made up of two flexible multilayers, one for the seat, the other for the backrest as for example Figures 4.1 and 4.2 detail it. FIG. 13 shows that the electrode layer 21 of this global multisensor is, at each millisecond, shared between 1 or more measuring electrodes R, peripheral electrodes at R which constitute the guard G and the rest in position E. The electrodes R and G are brought to zero potential, the electrodes E are brought to alternating potential V (fo). Therefore, it is tempting to simplify the structure of the multilayer sensor 10 (FIG. 2) by taking advantage of the stress measurements applied σ to the surface of the seat (therefore to layer 25 of the multilayer) by the passenger; and to measure the crushing of the layer 25 between the layer 21 and the second layer 21bis of the subsystem 2, but by measuring σ only in the regions of the seat and of the layer 21 which are, at the instant considered, in position E; therefore only the electrodes of 21bis placed opposite the electrodes E of 21 will be placed, sequentially, in the measurement position. And all the electrodes of 21bis placed opposite the electrodes R and G of 21 are placed in position G, so as to fulfill the essential role of rear guard for the electrodes R of 21. In other words, all the electrodes of 21bis are kept at the potential zero of G, referenced 56 (FIG. 14) except one of them or the group of adjacent electrodes of them which is placed in position R by the processor 58. Thus, at each cycle clocked by the processor 58, there are one or more electrodes grouped in position R or in the electrode layer 21, or in the electrode layer 21bis of the flexible circuit 41. The two sub-functions 1 and 2 are filled simultaneously. This preferred embodiment of the circuit 41 is thus reduced to three conductive layers: the layer of electrodes and their tracks 21, the layer of electrodes and their tracks 21bis and the layer 22 which is the rear guard of layer 21bιs. It is tempting to make the identical layers 21 and 21bιs; the electrodes are naturally symmetrical with respect to the median axis since the seat is itself symmetrical with respect to its median plane; the layers 21 and 21bιs can be identical, but mounted one at the place, the other upside down so that the tracks 51 of the electrodes 50 arrive on the connector 52 symmetrically. This provision is only of economic interest. Figures 15 illustrate this, 15.1 for the seat, 15.2 for the backrest.

The multisensor 41 of the seat can then be produced from an insulating film 23, PVC or other polymer, on which all the electrodes and their tracks are produced, this production has two modes: or else it is a method of depositing a conductive ink by screen printing or jet printing (jet-prmtmg), for producing the electrodes 50 and their tracks 51; or else the electrodes and their tracks are produced by laser ablation of the metallic layer uniformly deposited on the plastic film (alumized mylar or polypropylene, which are low-cost industrial products); and the ablation of metal is only carried out to trace the contour of the electrodes 50 and their tracks 51. The two production methods have been tested: the deposition of a flexible conductive ink is very economical but its lifespan is to be tested and the electrical conduction of the flexible ink, - although sufficient -, is much lower than that of the thin deposit (≤lμm) of aluminum of the above-mentioned product. The laser ablation (YAG) is perfect for the quality of the product produced, this method is also very adaptive in that it is easy to change the design of the electrodes and their tracks, by software The ablation of 0.3 μm of a-uminium deposited under vacuum on a film of polyester or other plastic does not require much power- a YAG laser succeeds very well, we were able to "write" the word "PICD" on a film of polypropylene alumized whose l 'thickness total is 15 μm, and where the ablation, by the YAG laser beam, of the aluminum has not attacked the plastic film; the mechanical robustness of this plastic film is unchanged after the laser ablation process which traced the letters. The laser ablation can be done with an ordinary plotter whose pen has been replaced by a fiber bringing the YAG laser beam to the focusing optics which replaces the pen. The cutting speed is 0.7m / s for a laser power of 2 Watt. Because the lines to be drawn to trace the contours of the electrodes 50 and their tracks 51 are very predominantly parallel respectively to the center axis of the seat (Y) and to its perpendicular (X), it is advantageous to multiply the number of beams UV laser and it is advantageous, to optimize the use of these "feathers", to draw the shape of the electrodes 50 and their tracks 51 so that several lasers can "evaporate" the metal simultaneously.

It should now be noted that tracks 51 are not electrically guarded on the front, passenger side

(below the seat), - they are kept on the rear face by the layer 21bis of which all the electrodes and tracks are at zero potential: guard potential of 56 for all, and the measurement potential 55 for the electrode (s) temporarily in measurement position. It is highly desirable that the tracks 51 are kept laterally from each other, this is why between each track 51 is placed a conductive tab permanently maintained at the guard potential 56. The only drawback of this interlacing of the tracks 51 and the tabs guard is that the laser ablation time is increased; but the measurements are significantly improved by these guard tabs between tracks, visible in Figure 15-3. It still remains to optimize the number of electrodes used in layer 21 for subsystem 1 and in layer 21bis for subsystem 2 according to the level of αuality desired for the PICD. Whatever the embodiment of the single-sided circuits (23 + 21) or (24 + 21bis), the PICD consists of the stacking of these two circuits placed face to face, with the layer of elastic dielectric 25 as an intermediate between the electrodes 50 of layers 12 and 14; between the part of the tracks 51 located between the electrodes 50 and the connector 52, a simple dielectric was placed, that is to say an ordinary plastic film; and of course there is a direct connection of the tracks and connection terminals 52. The conducting plane 22 which is the rear guard (potential 56) of the electrodes and tracks 21bis can be an independent metallized film or else, the plastic film 24 is metallized two sides and the laser ablations are only carried out on one side to produce the layer 21bis while leaving the layer 22 intact.

Figures 4 and 15 show an exemplary embodiment of the connector for the particular case of 48 and 24 electrodes respectively for the seat and back electrodes, the layers 21 and 21bis having exactly the same number of electrodes since it was chosen here to use identical layers 21 and 21bis, that is to say having, seen from above, the symmetry with respect to the median axis of the seat since the layers (23 + 21) and (23 + 21bis) are placed face to face on either side of 25. The 4 sides of the connector are occupied on 2 rows of terminals (144 terminals) by the tracks of the 4 single-sided circuits 21 and 21bis of the seat and the backrest. Figures 15 show in particular the guard tabs of the four layers 21, 2Ibis of 41 and 42 which separate the tracks to the connector 52 and the electrodes; we also see the reserved areas, without electrode or track, for the seams 2 in number for the backrest and only one for the seat. The choice shown in Figures 4 or 15 is not the most suitable for the easy discrimination between a passenger and a water or wine cubitainer placed on the seat: bypassing the tracks from the seat to the periphery of the circuits 21 and especially 21bis as shown for the file would be more judicious to distinguish the passenger from a cubitainer who, him, not having two buttocks, does not exert a stress on the seat significantly weaker in the center.

This example is actually a PICD specially made to test the relationship between the overall quality of passenger position detection and measurement and the effective number of electrodes, several of the 48 and 24 electrodes of each of the identical layers 21 and 21bis of the 'seat and back can be grouped inside the connector 52; these groups of adjacent electrodes are also generally different for 21 and for 21bis. To illustrate this, let us consider figure 16 which represents a single face circuit of 150x300mm, with its 12 electrodes whose unit surface is 29,2cm ² ; twelve electrodes or three electrodes if they are grouped into three sets of four electrodes or even six sets of 2 electrodes (figure lδbis); here the tracks of these electrodes are constituted by two cables, each of them with six conductors of 0.3mm in diameter under a braided sheath of external diameter 1.8mm, maintained at the guard potential 56; the six electrodes respectively connected to these six conductors can only be R or G on the one hand OR E on the other hand; it is indeed impossible to supply these six conductors under the same guard braid simultaneously at potential 57 (V ₀ (f ⁰ )) and at zero potential (55 or 56); but this restriction of freedom of choice is hardly significant and does not prevent the type of single-sided circuit 17 from achieving an acceptable version of 41, for example by using three circuits 17 for the file and four of these circuits 17 for 1 ' seated . The whiplash prevention device is now described. It is a simple adaptation of the PICD according to the invention, already announced previously. The principle of rear impact detection is the measurement of the crushing of the elastic layer 25, by the electrodes of the subsystem 2 located at the bottom of the backrest: the passenger during a rear impact undergoes a strong acceleration (20 to 40 g) essentially horizontal and which results in a bearing force on the lower part of the backrest equal to 20 or 40 times the usual weight of the passenger: the crushing of 25 at the passenger's center of gravity is therefore very strong and its detection very easy: the corresponding measurements almost exceed - immediately their normal level: a high threshold crossed is the signal of a rear impact.

The important thing is to be able to react very quickly, that is to say in a few milliseconds. However, the time of the exploration cycle of the various electrodes of subsystems 1 and 2 is measured in seconds. Hence the modification of the PICD with the addition of a second capacitive measurement bridge assigned only, and therefore permanently, to the measurement of the 2, or even 3, electrode (s) of the subsystem 2 located at the lower part of the file; detection of a rear impact is therefore acquired in 1 ms.

Immediately crossing the threshold of this signal induces the interruption of the current multiplexing sequence; 58 immediately causes the electrodes at the top of the backrest to be positioned in position R to determine the position of the passenger's head; if the head is resting on the headrest of the seat (the capacities measured are high), no action is triggered; if on the contrary, the capacities mentioned above are weak, it is because the passenger's head is placed far from the headrest and that the strong acceleration induced by the rear impact will cause the passenger's cervical vertebrae to rupture: the non-action threshold not being crossed, the processor 58 controls the firing of an airbag placed in the headrest or at the top of the backrest so as to fill the void between the head in ten milliseconds of the passenger and the top of the backrest and thus prevent the whiplash. It is desirable that the power of this airbag of a new kind be adapted to the distance between the head and the backrest, or that this airbag is fitted with several low, medium and high power cartridges to best fill, neither too much, nor too little, the volume between the passenger's head and his seat back. Figure 17 shows a version of the PICD also suitable for detecting a rear impact and preventing whiplash. FIG. 18 is the diagram of the preferred version of ASIC 40, in accordance with patent n ° 9613992, but modified here since two bridges of capacitive measurements are necessary: bridge 54 for subsystems 1 and 2 of the PICD and the bridge 64 operating for the sole electrodes at the bottom of the backrest (21bis) to permanently monitor the appearance of a rear impact and reacted in the millisecond, interrupting the multiplexing program of 53. These two bridges 54 and 64 have the same guard potential 56. All the electrodes at the bottom of the backrest are subtracted from multiplexing: those of 21 are kept permanently at potential 57 (always E) and those of 21bis, grouped together in a single electrode, are connected to the bridge input 64

(entry E2 in figure 18). The other electrodes 50 are connected to the inputs El which are the outputs of the multiplexer 53, which receives the potentials 55 (lug - from Al), 56, 57 here represented EPM = 0SC = 57. Bridges 54 and 64 have the same scheme.

The last part of the description relates to the replacement of the elastic film 25 by a set of air bags which are all parts of the comfort cushion announced previously. Each airbag 30, produced by heat sealing a tube or thin plastic sheets, to which is welded (or glued) a small diameter plastic tube 31, the other end of which is connected to a solenoid valve 32 (FIG. 20) . These airbags are placed between the single-sided circuits 21 and 21bis previously described. The solenoid valves 32 can inflate and deflate these bags under the control of the measurements of the PICD subsystem 2 for:

- or else maintain an almost constant distance between 21 and 21bis;

- or have the passenger's rear, legs and back wrapped like an anti-decubitus chair, therefore essentially thanks to a uniform pressure in the largest number of airbags, but with maintenance of the capacities measured by 21bis, therefore with bag volumes which correspond to the overall position that the passenger will have chosen himself and that 58 will have memorized, or even several positions taken successively (and memorized) and that 58 will pilot according to a defined sequence at selected time interval, see finally to modify, not statically, but dynamically, the shape of the seat, which induces the equivalent of a massage.

The bags 30 and the electrodes 21 can be microperforated (mechanically or by laser) to allow a certain flow of compressed air to leak (at ≡ 10-30 mbar) from the inflated bags; the advantage is to provide ventilation, therefore passenger comfort; the solenoid valves 32 then exclusively serve to inflate the bags.

In all cases, replacing the elastic film 25 with the series of inflated bags adds the comfort of largely eliminating the vibrations induced by the vehicle: vibrations transmitted by the engine or by the bearing via the tires; these cushions constitute a low pass filter which cuts all the high frequencies. The PICD is integral with the seat, therefore insensitive neither to the more or less advanced position of the seat in the passenger compartment, nor to the inclination given to it by the passaαer. The PICD system can not only be used for the purpose of inhibiting or authorizing the ignition of safety devices, but also to activate an anti-theft device: siren, inhibition of engine ignition, radio emission for locate the stolen vehicle, etc.

All of the above applies equally to a person's chair or bed.

Another application of the same method and / or system is the aid for automatic loading of containers, by detecting the positions of these containers, in a truck for example. The electrode mat is placed on the walls of the freighter, laterally or on the ceiling. The signal delivered by this system thus used gives the position and the volume of the containers.

Another application is the surveillance of a corridor or a room for residential or office use ... The electrode mat, placed under the floor or wall carpet, is an invisible and inviolable passage detector or presence of one or more people; their number and instantaneous positions are perfectly identifiable.

The system according to the invention also allows the recognition of shapes or parts circulating on a conveyor; this system also allows the metrology of a mechanical part of complex shape. The advantage of using the concept and / or the system according to the invention is the economy and the quality of this large number of capacitive measurements; economy, since only one capacitive measuring bridge is used; but also the quality, since one or more sensors can be replaced by as many standard capacities, which makes it possible to check, or even recalibrate, almost permanently, the zero, the gain, the thermal drift, etc., of the single capacitive measuring bridge used. This or these standard capacities have a first armature connected from time to time to the measurement potential Vm 55 and a second armature permanently connected to the injection potential Vi 57, and are surrounded on all sides, apart from the two potential inputs Vm and Vi, by a conductor permanently brought to the guard potential Vg 56.

If the part to be metrologically checked is insulating, all the electrodes (the n of R being measured and the p of G on the one hand, the Nnp of E on the other hand) electrostatically influence this part and fix its alternating potential at an intermediate value between V ₀ (f °) and zero; metrological control is really only valid for measuring deformations.

On the other hand, if this part is metallic (or only slightly conductive of electricity), if it can be isolated from the Earth and if one can apply to it directly - and not by electrostatic influence - the potential V ₀ (f °), then it is a true metrology, absolute, and no longer only relative of measurement of deformations from an initial state, that it is possible to achieve. In addition, the network of switches 53 is simplified, since it is reduced to a single switch 531 for the two switched functions: that of measurement by, in general, only one electrode at a time and electrical guard for all the others. Metrology of essentially cylindrical objects, rods or bores, smooth or threaded, even grooved (gears) or striped

(arming tubes) is particularly advantageous by a measurement system according to this particular form of the invention. The printed circuit 41, for example made of very flexible double-sided capton, is simply glued to a solid non-deformable skeleton; it does not matter that the geometry of the electrodes placed on the front face of this circuit is fairly poor and even unstable over time, in particular as a result of the action of atmospheric humidity on the plastic material of the double-sided circuit; the measurement of a well-identified standard object, or even certified by an authorized body, makes it possible to know, with an accuracy almost equal to that of the standard, the exact geometry of the multisensor at the time of this calibration on the standard; and therefore to measure the cylindrical part to be checked with very high accuracy and speed of metrology.

Of course, the invention is not limited to the examples which have just been described and numerous modifications can be made to these examples without departing from the scope of the invention. For example, provision may be made for the stress detection subsystem to comprise means for carrying out resistive detection, these resistive detection means comprising a conductive rubber strip having an elastic-conductive effect.

Claims

RESENDTCATTONS

1. Capacitive measurement system, for detecting a presence and identifying objects or people in a given environment, comprising a set of electrodes, multiplexer means (53) for connecting these electrodes to a capacitive measurement electronics (54), control and processing means (58) for controlling the multiplexer means (53), acquiring and processing the measurement results, this set of electrodes (50) consisting of one or more groups of electrodes (41, 42 ) each comprising several electrodes (50) at least in part substantially juxtaposed and contiguous to form as many pads of electrodes, characterized in that the control and processing means (58) and the multiplexer means (53) cooperate to assign to each electrode (50) has any function among the respective functions for measuring, guarding and injecting charges, said electrodes (50) having no predefined and immutable function.

2. System according to claim 1, with a capacitive measurement electronics (54) comprising a charge amplifier whose negative input (55) at a measurement potential Vm is connected to the measurement electrodes and whose positive input (56 ) is connected to the guard at a guard potential Vg equal to the measurement potential Vm, and a source of potential Vi (57) such that Vm-Vi = Vo (fo), Vo being the amplitude of an alternating voltage of frequency fo, characterized in that the multiplexer means (53) are controlled by the control and processing means (58) to apply to each electrode (50) a potential among the three respective measurement potentials Vm, guard Vg and d injection Vi.

3. System according to claim 2, characterized in that the measurement potential and the guard potential are substantially zero, that is to say taken as a potential reference and corresponding in particular to the potential of a vehicle body.

4. System according to one of claims 2 or 3, characterized in that the electrode mat (41) is a multilayer circuit comprising at least a first flexible plastic film (11) comprising on one of its faces a layer (12) situated on the side of the target to be detected and making the capacitive measurement electrodes and their tracks for connection to the electronics (40), and a second flexible plastic film (11) comprising on one of its faces a fully metallized layer (15) acting as a rear guard for said capacitive measurement electrodes.

5. System according to claim 4, characterized in that the multilayer circuit further comprises a third flexible plastic film (11) having on one of its faces a conductive heating layer (16).

6. System according to claims 3 and 4, characterized in that the electrode mat (41) comprises a single layer obtained by screen printing, or chemical attack or laser attack, and covered with a varnish.

7. System according to any one of the preceding claims, comprising a capacitive identification subsystem comprising a set of electrodes, multiplexer means (53) for connecting these electrodes to a capacitive measurement electronics (54), means control and processing (58) to control the multiplexer means (53), acquire and process the measurement results, this set of electrodes (50) consisting of one or more groups of electrodes (41, 42) each comprising several electrodes (50) at least partly substantially juxtaposed and contiguous to form as many electrode pads, characterized in that it further comprises a subsystem for detecting mechanical stresses applied to these electrode pads.

8. System according to claim 7, in which the capacitive identification subsystem comprises at least one flexible multilayer printed circuit containing the electrode pads, characterized in that the stress detection subsystem comprises means for measuring variations in capacity between two adjacent layers of the printed circuit.

9. System according to claim 8, characterized in that the flexible multilayer printed circuit comprises: a first layer (12) receiving the electrodes of the identification subsystem, - - a second layer (13) constituting an electrical guard plane first layer electrodes; - a third layer (14) receiving capacitive measurement electrodes for the applied stresses; and - a fourth layer (15) brought to an alternating potential.

10. System according to any one of claims 8 or 9, characterized in that the flexible multilayer printed circuit is produced by superposition of a first flexible film (21) metallized on one side, of a second flexible film (25) dielectric, and a third flexible film (21bis) metallized on its two faces, these three films being bonded together.

11. System according to claims 8 and 10, characterized in that the two adjacent layers are constituted by the first (21) and the third (2lbis) flexible films separated by the second elastic dielectric flexible film (25) which reacts to the applied stress.

12. System according to any one of claims 8 to 11, characterized in that the flexible multilayer printed circuit is housed in a vehicle seat, in particular between the body of the seat and the covering covering of this seat.

13. System according to claim 12, characterized in that the flexible multilayer printed circuit comprises empty strips of electrode reserved for seams of the covering covering of the seat.

14. System according to claim 7, characterized in that the stress detection subsystem comprises means for measuring an air pressure in elementary airbags for a predefined crushing of each of said airbags.

15. The system of claim 14, characterized in that the predefined crushing of each of said airbags is measured capacitively.

16. System according to one of claims 14 or 15, implemented in a seat, characterized in that the airbags are independent and adjacent and constitute an assembly forming an airbag arranged in the seat and in the back of said seat .

17. System according to any one of claims 14 to 16, characterized in that the airbags are connected to solenoid valve means controlled by the control and treatment means for inflating or deflating said airbags.

18. The system of claim 17, characterized in that the solenoid valve means are programmed to selectively inflate some of said airbags so as to wrap a determined part of the body of a person placed on a support structure provided with said system.

19. The system of claim 18, characterized in that it is further arranged to store personalized selective inflation states.

20. System according to any one of claims 17 to 19, characterized in that the solenoid valve means are controlled to effect selective dynamic inflation of the airbags so as to provide a massage function.

21. Presence detection system according to claim 7, characterized in that the stress detection subsystem comprises means for performing resistive detection, these resistive detection means comprising a conductive rubber strip having an elastic-conductive effect .

22. System according to any one of claims 7 to 21, characterized in that the stress detection subsystem is arranged to detect a rear impact on the vehicle, by measuring the pressure exerted by a passenger on the bottom of the backrest. .

23. The system of claim 22, characterized in that the capacitive identification subsystem is immediately arranged to detect the position of the passenger's head relative to the backrest, as soon as the upper threshold for measuring the bearing force of the passenger on the bottom of the backrest is exceeded.

24. The system of claim 23, characterized in that it is arranged to control the triggering of an air cushion device disposed in the top of the seat to retain the head of the passenger, in the event of detection of a position too away from the head in relation to the file.

25. A method for detecting the presence and identifying people or objects in a given environment, implemented in the system according to any one of the preceding claims, comprising:

- capacitive measurement sequences carried out by the electrode pad (s) (41, 42),

- sequences for processing said measurements in order to extract presence detection and identification information therefrom, characterized in that, for each measurement sequence on a carpet, the set of electrodes constituting is defined dynamically this mat: - a field (I, E) of electrodes which will be subjected to the injection or influence potential,

- a field (M, R) of electrodes which will be subjected to the measurement potential, and

- A domain (G) of electrodes which will be subjected to the guard potential, this guard domain (G) being chosen so that it best surrounds the measurement domain (R).

26. The method as claimed in claim 25, characterized in that the measurement domain (M) forms a compact block of adjacent electrodes, and in that the guard domain (G) is formed by at least all of the first electrodes neighboring the electrodes forming the measuring range (M).

27 Method according to one of claims 25 or 26, characterized in that the measurement processing sequences use a neural network method

28. Method according to any one of claims 25 to 27, implemented in the system according to any one of claims 7 to 24, comprising a capacitive identification of a passenger using at least one mat of electrodes housed in a seat, characterized in that it also comprises a detection of mechanical stresses applied to this carpet of electrodes.

29. Application of the system and method according to any one of the preceding claims, for controlling the activation of safety devices in a motor vehicle, in particular airbag type devices or anti-theft devices, characterized in that a or several pads of electrodes are arranged under the covering of the seats and / or the interior walls of said motor vehicle.

30. Application of the system and method according to one of claims 1 to 28, for detecting and measuring the position of containers (containers) within a transport system, in particular a truck or a cargo ship, characterized in that a or several pads of electrodes are arranged on one or more interior walls of said transport system.

31. Application of the system and method according to one of claims 1 to 28 to the metrology of mechanical parts, characterized in that one or more measuring electrodes are replaced by as many standard capacitors having a first time-connected armature another to the measurement potential Vm and a second armature permanently connected to the injection potential Vi, said standard capacitor (s) being surrounded on all sides, apart from the two potential inputs Vm, Vi, by a conductor permanently carried to guard potential Vg.

32. Application of the system and method according to one of claims 1 to 28, for detecting and measuring the presence of intruders in a room, characterized in that one or more pads of electrodes are arranged on one or more walls interior of said room.

33. Application of the system according to one of claims 13 to 20, to detect the presence of a person placed standing in front of a chair and activate the movement of the seat of this chair to assist the person during the phase allowing move from standing to sitting position.