Bias-corrected accelerometer-type cold atom sensor

Abstract

The invention relates to an ultracold atom inertial sensor (10) of accelerometer type comprising: an atom chip (ACh) comprising at least one set of a first and a second elementary sensor (SENA, SENB), an atom generation device (ACG), a generator (GB) of a homogeneous magnetic field (Bc), a power supply device (PSD), the arrangement of said group of one or more conductive elements of each sensor and said sequence being further configured so that the trajectories (TR1A, TR2A) associated with the first elementary sensor and the trajectories (TR1B, TR2B) associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously and so that the departure direction of the first clouds, respectively of the first and second elementary sensors, are opposite, the sensor further comprising a detection system (SDET). Figure to be published: Figure 7.

Description

Title of the invention: Cold atom sensor of the bias-corrected accelerometer type FIELD OF THE INVENTION

[0001] The present invention lies in the field of inertial sensors and more specifically cold atom interferometric inertial sensors integrated on an atomic chip. More particularly, the invention relates to cold atom sensors on an atomic chip of the accelerometer type using microwave fields for the spatial separation of the two clouds of atoms according to two internal states, used when measuring the inertial parameter acceleration. STATE OF THE ART

[0002] Generally speaking, inertial sensors are devices for measuring physical parameters related to movement such as accelerations or angular speeds, which, combined with a clock, allow one to locate oneself in space. A device measuring acceleration and rotational speed is called an inertial unit or inertial measurement unit (IMU).

[0003] On-chip cold atom interferometry technology makes it possible to perform this type of inertial measurements, and tends to demonstrate sufficient compactness and performance for its use in embedded fields.

[0004] The operation of the cold atom accelerometer is based on interferometry, and the measurement of the phase difference of the interferometer makes it possible to trace back to a measurement of the acceleration.

[0005] A cold atom interferometer interferes two electronic states, called first internal state 1a> and second internal state 1b>, of an atom, such as rubidium 87, in a Ramsey-type sequence. A Ramsey-type interferometer sequence measures a phase q> accumulated during the implementation of the sequence, from a measurement of at least one population of a chosen state 1a> or 1b> (preferably a measurement of both populations for greater precision).

[0006] To make this interferometer sensitive to accelerations, the two states must both travel a rectilinear trajectory back and forth starting from the same starting point DP, the two trajectories and the starting point being aligned along a straight line Dr. Thus, a term dependent on the difference in energy acceleration potential between the positions of the two states 1a> and 1b> is added to the energy difference between the two states and therefore the phase of the interferometer depends on the acceleration.

[0007] The operating principle of measuring acceleration by interferometry is briefly recalled below, based on a clock-type Ramsey sequence, where the phase of the interferometer at the end of the sequence is proportional to the energy difference of the two states of the interferometer.

[0008] In the case of a clock, the phase of the interferometer is given by:

[0009] m = <p - <p= {œ- m b) TR clock a + b x «

[0010] With:

[0011] is the phase of the atoms in the la> state, and ^b that of the atoms in the b> state,

[0012]

[0013] is the energy of state la> and is the energy of state lb>

[0014] is the pulsation corresponding to the energy difference between the states la> and lb>,

[0015] œ is the pulsation of the local oscillator,

[0016] TR is the free evolution time (Ramsey time).

[0017] If during this sequence, we make sure that the atoms in the state la> and those in the state lb> each describe a rectilinear back-and-forth trajectory as described previously, a phase shift term sensitive to the acceleration along the line Dr appears due to the difference in acceleration potential energy between the positions of the two states la> and lb>: the clock becomes an accelerometer.

[0018] If the separation is done along the X axis (straight line Dr along X), and an acceleration ac is present during the duration T^p during which the states are separated, then an additional phase shift appears at the output of the interferometer:

[0019] %.ce= -Xh(t))dt

[0020] With:

[0021] xa(t) is the position of state la> along line Dr and xb(t) is the position of state lb> along line Dr,

[0022] m is the mass of an atom and 72 is the reduced Planck constant.

[0023] Indeed, the energy levels of the two states of the interferometer have been shifted in energy. This shift in energy levels is due to the spatial separation of the two states; due to their spatial separation, the two states are no longer at the same level of potential acceleration energy.

[0024] To simplify the equation, we consider separation and recombination durations that are negligible compared to the duration where separation is maintained (Tsep * TXa and xb are therefore no longer a function of time and we can rewrite the previous equation in the form:

[0025] m with Ax = xa-xh vacc nx

[0026] The total phase at the output of the interferometer is therefore: 100271 = ^-(^+^)1^(1)

[0028] Knowledge of this phase Pfot as well as of the other quantities and parameters of formula (1) allows us to go back to the acceleration along the X axis.

[0029] The operating principle of an ultra-cold atom accelerometer based on an atomic chip is recalled below, which makes it possible to generate the clouds of ultra-cold atoms and the trajectories allowing the implementation of the measurement of the acceleration via the phase as described above.

[0030] Ultracold atoms are defined as atoms whose temperature is less than 400 nanokelvins, preferably 300 nanokelvins. The temperature of thermal ultracold atoms is, for example for Rubidium atoms, between 50 and 400 nanokelvins and preferably between 100 and 300 nanokelvins.

[0031] The principle is to produce two straight and aligned back-and-forth trajectories traveled respectively by the two clouds of magnetically trapped atoms. The movement of the magnetic trap along the trajectories is achieved by conductive wires / elements and microwave guides arranged on and in the atomic chip.

[0032] An example of an atomic chip is illustrated [Fig.l].

[0033] The surface of the chip defines an XY plane or measurement plane 13, normal to a Z axis.

[0034] The chip 1 comprises means adapted to generate a first ultracold atom trap T1 and a second ultracold atom trap T2, a trap making it possible to immobilize a cloud of ultracold atoms in an internal state different from the other trap, at a predetermined distance h from the measurement plane 13. For example, the trap T1 comprises the atoms in the electronic level or state la> (cloud CL1) and the trap T2 comprises atoms in the state lb> (cloud CL2). The levels la> and lb> are spaced apart by a frequency coab / 2ir. For example, in the case of rubidium 87, these are the two hyperfine levels IF=1,mF=-1> and IF=2,mF=1>, spaced apart by approximately 6.8 GHz.

[0035] These means also make it possible to move the clouds along the trajectories TRI and TR2 located in a plane parallel to the measurement plane 13, at a height h from this plane, as illustrated [Fig.l]. These means consist of waveguides and conductive wires.

[0036] The waveguides CPW1 and CPW2 are suitable for the propagation of microwaves at cob and coa pulses. The waveguides are arranged symmetrically with respect to a Y axis of the measurement plane, preferably parallel. The two waveguides CPW1 and CPW2 are connected to at least one microwave frequency voltage or current generator. For example, each of the waveguides is produced by depositing three parallel conductive wires to produce a coplanar waveguide. In other embodiments, other types of waveguides can be used, in particular waveguides whose manufacture is compatible with microfabrication techniques by deposition or by etching. For example, a microstrip line can be produced.

[0037] The conductive wires integrated into the chip 1 are adapted to be crossed by currents continuous. The conductive wires are distributed for example into a conductive wire WIz along an axis of symmetry Y perpendicular to X and included in the measurement plane 13, and into a conductive wire Wld, parallel to the X axis. The wires are arranged so as to define a crossing point C (crossing between WIz and Wld) located on the Y axis.

[0038] Each conductive wire is connected to one or more current and / or voltage generators, themselves connected to a processing unit comprising at least one microprocessor. The voltage and / or current generators make it possible to control both direct currents and alternating currents in the wires. In particular, direct currents are controlled in the conductive wires.

[0039] In the sensor, the atom chip 1 is placed in a vacuum chamber maintained, for example, using an ion pump and preferably comprising magnetic shielding. The sensor comprises a device for generating ultracold atoms which comprises:

[0040] - an atom emitter (dispenser in English), for example produced by a heating filament delivering a rubidium vapor;

[0041] -a primary atom trap (optical and / or magnetic), making it possible to pre-cool and arrange a cloud of ultracold atoms in the vicinity of the chip, to load the magnetic traps T1 and T2 described later with atoms.

[0042] The sensor also comprises a magnetic field source, external to the chip 1. It makes it possible to impose a homogeneous and stationary magnetic field Bc over a thickness at least of the order of a height h above the measurement plane 13. Advantageously, the direction of the homogeneous magnetic field is parallel to the measurement plane.

[0043] In [Fig. 1] the dotted trajectories TRI and TR2 illustrate respectively the trajectory of the ultracold atom clouds CL1 and CL2. A distance h separates the plane of the trajectories and the measurement plane 13 of the chip. Preferably h is between 500 nm and 1 mm, and preferably between 5 pm and 500 pm.

[0044] [Fig.2] illustrates the geometry of the guides and wires of the atomic chip as well as the traps Tl and T2. The specific arrangement of the conductive wires and waveguides, associated with the homogeneous magnetic field source, makes it possible to easily obtain two traps Tl and T2 as illustrated in part a) of [Fig.2]. Each trap Tl and T2 has a non-zero and identical minimum value V0, and an identical curvature, a necessary condition for the sensor to work. Indeed, when a direct current is applied to the two conductive wires crossing at C, the minimum potential is located vertically above this crossing point. When then a microwave power is sent into the waveguides, the central minimum is transformed into two minima on either side of the initial minimum in the direction of the waveguides. If the initial minimum is not located strictly at an equal distance from the two waveguides, the two potential minima created will not have rigorously the same value of the minimum VO and the same curvature.

[0045] Part c) of [Fig.2] illustrates the arrangement of the conductive wires defining the initial crossing point C and the waveguides (top view). Part b) of [Fig.2] describes the corresponding arrangement of the conductive wires and the waveguides printed on a chip in profile view, in section along the conductive wire Wld which crosses the conductive wire WIz along the axis of symmetry Y. The waveguides CPW1 and CPW2 are coplanar waveguides located on a first level NI. The insulating layer 18 advantageously makes it possible to flatten the measurement plane. The material of the electrical insulating layer may be, for example, silicon dioxide, silicon nitride or benzocyclobutene. A conductive material is used for the manufacture of the conductive wires, for example gold, and is deposited on a substrate 15, forming a second level N2. The substrate may be, for example, made of silicon, aluminum nitride or silicon carbide.

[0046] In part a) we see the symmetrical separation of ultracold atoms, specific to the internal state of the ultracold atoms, and more precisely the variations in potentials as a function of the X axis of chip 1.

[0047] Curve "a" shows a potential well corresponding to the association of the homogeneous magnetic field and the field created by the two intersecting conductive wires, the wire WIz carried by the current Iz and the wire Wld carried by the current Id. This results in a local potential well, initial potential Vini forming an atomic trap T in three dimensions. A cloud of ultracold atoms can be trapped and cooled there.

[0048] Curve "b" schematically shows the potential created by the transmission of microwaves at the cob frequency in the waveguide CPW1. The field emitted by the passage of microwaves at the cob frequency makes it possible to modify the energy of the ultracold atoms and to move the atoms of internal states lb>. Curve "e" illustrates the potential seen by the internal states lb> due to the contributions of the potentials illustrated by curve "a" and by curve "b". Curve "e" shows a local minimum of potential making it possible to locally trap a cloud of ultracold atoms of internal states lb>.

[0049] Similarly, curve "d" schematically shows the potential created by the transmission of microwaves at frequency coa in the waveguide CPW2. The field emitted by the passage of microwaves at frequency coa allows the energy of the ultracold atoms to be modified and the atoms of internal states la> to be displaced. Curve "c" illustrates the potential seen by the atoms of internal states la> due to the contributions of the potentials illustrated by curve "a" and by curve "d". Curve "c" shows a local energy minimum allowing a cloud of ultracold atoms of internal states la> to be locally trapped.

[0050] The combination of a DC magnetic trap (created by the direct currents in the wires and the homogeneous field Bc) and a microwave field creates what is called a "dressed" trap. By "dressed" is meant a trap created at least in part by an oscillating microwave, radio-frequency or optical field. Changes in the microwave fields (power, frequency and guide in which they propagate) make it possible to move this dressed trap and therefore to move the atoms. The DC magnetic trap is represented in [Fig.2] by curve a. The microwave field at coa is represented in [Fig.2] by curve d and the microwave field at cob is represented in [Fig.2] by curve b. The dressed trap Tl (combination of curves a and d) for the state la> is represented by curve c and the dressed trap T2 (combination of curves a and b) for the state lb> is represented by curve e.

[0051] The ultracold atom clouds of internal states 1a> and 1b> can be separated and trapped symmetrically with respect to the Y symmetry axis by simultaneously imposing the propagation of waves of frequency coa in CPW2 and cob in CPW1. To obtain two traps whose minima are of the same value V0 and whose curvatures are of the same value, it is important that the crossing point C is arranged at an equal distance from CPW1 and CPW2, on the Y symmetry axis.

[0052] The frequencies coa and cob are chosen as a function of the frequencies of the states la> and lb> cüo a and co0b.

[0053] [Fig.3] illustrates in its lower part b) an example of application of the different signals to the wires and guides according to the state of the art to make travel, in the case of an accelerometer type sensor, the linear trajectories to the two traps T1 and T2 (and therefore to the two trapped atom clouds), and in its upper part a) the position of the two clouds corresponding to chosen times. Thus part a) of [Fig.3] schematically presents a sequence of the displacement of each of the ultracold atom clouds at characteristic times ti to t3. Part b) illustrates the sequence of the different currents applied to the conductive wires, of the powers applied to the waveguides and of the frequencies imposed on the waveguides, for the times corresponding to those of part a). In the sequence presented in [Fig.3], the current Iz, not shown, circulating in WIz is stationary, at a constant value.The current applied to Wld is also constant throughout the sequence.

[0054] In a step A0, there is a phase of preparation of the atoms. A cloud of ultracold atoms is generated, including phases of emission of said atoms, of cooling of said atoms, of initialization of said atoms in at least one internal state 1a> and of trapping of a cloud of said ultracold atoms in a local minimum of potential, at a distance h from the measurement plane (trap T, curve “a” of [Fig.2] part a)). The height h is different from 0 because the homogeneous magnetic field Bc is non-zero. The trapping is carried out by applying direct currents in the wire WIz and in the wire Wld (Vini generation). At the same time, a bias magnetic field Bc is applied parallel to the plane of the atom chip, which is superimposed on the magnetic field created by the two previous wires. The atom cloud is then trapped by the Vini potential vertically above C, the intersection of the Wlz and Wld wires. This step is illustrated by the part between 0 and ti of [Fig.3] part b).

[0055] In a step B0 the internal states are initialized by coherently superimposing the ultracold atoms between the states 1a> and 1b> by a first pulse TT / 2 . This pulse can be produced by a laser, microwave emission, or more generally by a method emitting waves at a suitable transition frequency. The currents Iz and Id are imposed respectively on the conductive wires Wlz and Wld. The two internal states 1a> and 1b> are coherently and spatially superimposed in line with the crossing point C.

[0056] In a step CO, the initial cloud is spatially separated into a first cloud of ultracold atoms CL1 in a first internal state la> and a second cloud of ultracold atoms CL2 in a second internal state lb>, forming a first ultracold atom trap Tl and a second trap T2. This step is illustrated by the part between ti and t'i of [Fig.3] part b).

[0057] In a step D0, the separation of the two traps is maintained for a time Ts (t'i to t'2) then the traps are recombined at said starting point (t'2 to t3).

[0058] During steps C0 and D0 the two clouds respectively travel a first linear trajectory TRI and a second linear trajectory TR2 parallel to the XY plane from the starting point DP.

[0059] In a step E0 the internal states 1a> and 1b> are recombined by applying a second rr / 2 pulse to the ultracold atoms, which transfers the phase difference to the populations of the two atomic levels. The 7T / 2 pulses can be sent to the atoms via the microwave guides or via a separate microwave transmitter.

[0060] The sequence from the first to the second pulse æ / 2 inclusive is the Ramsey sequence (see above).

[0061] Then in a step F0 we measure the phase due to the difference in acceleration potential energy between the two states 1a> and 1b> of the ultracold atoms and we calculate the acceleration of the sensor along the X axis.

[0062] Considering only the Ramsey sequence of duration TR between ti and t3, initially at ti no power is applied to the waveguides and the cloud is trapped above point C. Between ti and t'i, the microwave power injected into the waveguides CPW1 and CPW2, gradually increases from 0 to its maximum value, then the value remains maximum and constant between t'i and t'2 passing through t2. A coa pulse is sent into the waveguide CPW2 and a cob pulse is sent into the waveguide waveguide CPW1, which makes it possible to separate the two clouds of different internal states on either side of the Y symmetry axis up to the positions shown in t2, these positions being maintained for the entire duration t'2-t' b which can be small. Between t'2 and t3, the microwave power in the waveguides is gradually cut off. The two traps T1 and T2 move until they merge into a single trap located at the starting point shown in t3 (part a)).

[0063] In order to implement the method described above, the ultracold atom sensor allowing measurement of the acceleration ac comprises:

[0064] -an atomic chip 1 as described previously, with the waveguides and the conductive wires,

[0065] -an atom generation device for generating the ultracold atom cloud near the measurement plane 13 of the atom chip,

[0066] -a generator of the homogeneous magnetic field Bc

[0067] -at least one direct voltage or current generator adapted to control electric currents in the conductive wires and at least one microwave voltage or current generator connected to the waveguides,

[0068] - a detection system, typically of an optical intensity, adapted to measure at least one population of ultracold atoms in an internal state, this measurement allowing the determination of the phase due to the difference in acceleration potential energy between the two states la> and lb> and therefore the determination of the acceleration ac.

[0069] To obtain a more precise measurement of the inertial parameters (time, acceleration, rotation speed) it is possible to use measurement redundancy. Architectures integrating several sensors, clocks, accelerometers and gyrometers are described in document US2023 / 0178262. The sensor is based on an AchMO matrix chip allowing the production of several elementary sensors.

[0070] [Fig.4] illustrates an example of a 6x6 matrix of this document. The waveguides along 6 axes Xn and along 6 axes Ym form the columns and rows of the matrix. Each pixel (n,m) of the matrix corresponds to a potential elementary sensor that can be configured according to the need. For example, the chips in column C1 measure the acceleration along X ax, the chips in column C2 measure the rotation speed around the Y axis Qy, the chips in row L1 measure the acceleration along Y ay and the chips in row L2 measure the rotation speed around the X axis Qx. As a measurement requires a particular sequence for the coplanar guides, these cannot be shared for two simultaneous measurements of two distinct inertial parameters. Thus the pixels 4 surrounded by a circle are not used.The matrix chip is thus reconfigured according to the needs: the type of measurement desired (ax, ay, Qx, Qy, t), the desired precision (function of the number of chips performing simultaneously. the measurement), etc. Parallel, redundant and / or complementary measurements are thus carried out on the same matrix chip. Note that a measurement of Qz is also possible with a particular sequence of application of direct currents and / or the magnetic field and / or microwave fields, with an adapted chip topology if necessary (see for example documents US2023 / 0178262 and US2022 / 0397396).

[0071] [Fig.4bis] illustrates another example of a 6x6 matrix of document US2023 / 0178262 in which the waveguides form the diagonals, numbered (DI to D6, then Dl' to ...). The rows and columns are formed by the DC wires Win and W2m. Pixels of the South East-North West diagonal D5 measure ax, some pixels of the diagonal D6 measure Qy. The South West-North East diagonal D6' measures Qx, and the South West-North East diagonal D7' measures ay. Thus the elementary chips 5 surrounded by a circle are not used for the same reasons as previously.

[0072] Sensors based on a matrix chip comprising a plurality of elementary sensors according to the various geometries described above make it possible to:

[0073] Carry out on a single chip the clock function, the measurement of the acceleration along two orthogonal axes and the measurement of the rotations along three orthogonal axes two by two,

[0074] Duplicate the elementary geometry on the same atomic chip to carry out all the previously mentioned measurements in parallel and implement redundancy between the different measurements.

[0075] However, this solution does not allow the elimination of measurement biases or error correction.

[0076] Every measurement is associated with an uncertainty that results in an error in the measurement. This error can be due to random factors that are unavoidable and cannot be corrected (however, their amplitude can be estimated), but also to systematic factors due to imperfections in the sensor or imperfect measurement conditions. It is this last type of error that is also called bias.

[0077] An aim of the present invention is to remedy the aforementioned drawbacks by proposing an accelerometer-type inertial sensor and a measurement method allowing the correction of certain biases, in order to obtain better performance of the sensor. DESCRIPTION OF THE INVENTION

[0078] The present invention relates to an inertial sensor with ultracold atoms of the accelerometer type comprising: • an atomic chip placed in a vacuum chamber, comprising an XY plane called the measurement plane normal to a Z axis, and comprising at least one set of a first and a second elementary sensor, each elementary sensor comprising: • at least a first pair of waveguides parallel to each other and at least a second pair of waveguides parallel to each other and intersecting with the first pair, • a group of one or more conductive elements, an atom generation device configured to generate an initial cloud of ultracold atoms associated with the first elementary sensor and an initial cloud of ultracold atoms associated with the second elementary sensor, said initial clouds being located near said XY plane of said chip, a generator of a homogeneous magnetic field (Bc), a power supply device comprising at least one microwave generator and at least one direct current generator, the power supply device being configured to apply microwave signals to said waveguides and direct currents to said conductive elements, the magnetic field generator and the power supply device being configured to apply the magnetic field, the direct currents and the microwave signals in a predetermined sequence, an arrangement of said group of one or more conductive elements and said sequence being configured, upon implementation of each sensor, to: i. Generate an initial trapping potential of said initial cloud of ultracold atoms defining a starting point, ii. Spatially separating the initial cloud into a first cloud of ultracold atoms in a first internal state from a second cloud of ultracold atoms in a second internal state, respectively forming first and second ultracold atom traps, and such that the first cloud, the second cloud and the starting point are aligned along a measurement axis, iii. Maintaining the separation of said first and second traps for a time Ts and then recombining said traps at said starting point, so that said first and second clouds respectively travel a first and a second linear trajectory parallel to the XY plane from the starting point, the arrangement of said group of one or more conductive elements of each sensor and said sequence being further configured so that the trajectories associated with the first elementary sensor and the trajectories associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously and so that the starting direction of the first clouds, respectively of the first and second elementary sensors, are opposite, • the sensor further comprising a detection system adapted to measure at least a first phase of the first elementary sensor and a second phase of the second elementary sensor, an acceleration being determined from a difference between the first and second phases.

[0079] According to one embodiment, the power supply device PSD is further configured to apply direct currents to said waveguides.

[0080] According to one embodiment, the first and second pairs of waveguides are perpendicular to each other, and the sensor comprises at least two assemblies configured to carry out an acceleration measurement respectively along two orthogonal axes.

[0081] According to one embodiment, the atomic chip has a matrix structure whose pixels define potential elementary sensors, a set comprising two pixels of the matrix.

[0082] According to one embodiment, the sets of a first and a second elementary sensor have at least one conductive element in common.

[0083] According to another aspect, the invention relates to an inertial unit comprising at least two accelerometer-type sensors according to the invention configured to carry out an acceleration measurement respectively along two orthogonal axes, the atomic matrix chip further comprising pixels configured to carry out at least one clock measurement and pixels configured to carry out rotation speed measurements along three orthogonal axes.

[0084] According to another aspect, the invention also relates to a method for measuring an acceleration by an interferometric ultracold atom inertial sensor comprising an atomic chip placed in a vacuum chamber, comprising an XY plane called the measurement plane normal to a Z axis, and comprising at least one set of a first and a second elementary sensor, each elementary sensor comprising: • at least a first pair of waveguides parallel to each other and at least a second pair of waveguides parallel to each other and intersecting with the first pair, • a group of one or more conductive elements, the method comprising the steps of: A. Generating an initial cloud of ultracold atoms associated with the first elementary sensor and an initial cloud of ultracold atoms associated with the second elementary sensor, said initial clouds being located near said XY plane of said chip, B. Generate a homogeneous magnetic field, C. For each sensor generate an initial trapping potential of said initial cloud of ultracold atoms, D. For each sensor initialize a first internal state and a second internal state by a first pulse ir / 2, E. For each sensor, spatially separate the initial cloud into a first cloud of ultracold atoms in the first internal state from a second cloud of ultracold atoms in the second internal state, respectively forming a first and second ultracold atom trap, and such that the first cloud, the second cloud and the starting point are aligned along a measurement axis, F. For each sensor, maintaining the separation of said first and second traps for a time Ts then recombining said traps at said starting point, so that said first and second clouds respectively travel a first and a second linear trajectory parallel to XY from the starting point, steps B to F being carried out by applying, according to a predetermined sequence, a homogeneous magnetic field, direct currents to said conductive elements and microwave signals (IMw) to said waveguides,

[0085] the arrangement of said group of elements of one or more conductive elements and said predetermined sequence being further configured so that the trajectories associated with the first elementary sensor and the trajectories associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously, and so that the departure direction of the first clouds, respectively of the first and second sensors, are opposite, A. Recombine said first and second internal states by applying to said ultracold atoms a second pulse ji / 2, B. Measure at least a first phase of the first elementary sensor and a second phase of the second elementary sensor, an acceleration being determined from a difference between the first and second phases.

[0086] According to one embodiment, step C comprises the application of direct currents in at least one conductive element.

[0087] According to one embodiment, step E and step F comprise the application of microwave signals in a pair of waveguides in order to measure an acceleration along a measurement axis perpendicular to the waveguides of said pair.

[0088] According to one embodiment in step E and step F the application of microwave signals comprises: • for the first elementary sensor the application of a signal of frequency coa in one of the waveguides and the application of a frequency cob in the other waveguide, • for the second sensor the application of a cob frequency signal in a waveguides and applying a coa frequency in the other waveguide, said frequencies coa and cob being determined from the resonance frequencies of said first and second internal states.

[0089] According to one embodiment, step F comprises a sub-step of cutting off the microwave signals and a sub-step of applying direct currents to said waveguides.

[0090] The following description presents several exemplary embodiments of the device of the invention: these examples are not limiting of the scope of the invention. These exemplary embodiments present both the essential characteristics of the invention as well as additional characteristics linked to the embodiments considered.

[0091] The invention will be better understood and other characteristics, aims and advantages thereof will appear during the detailed description which follows and with reference to the appended drawings given as non-limiting examples and in which:

[0092] [Fig.l] illustrates an example of an atomic chip according to the state of the art.

[0093] [Fig.2] illustrates the geometry of the guides and wires of an atomic chip according to the state of the technique, as well as magnetic traps.

[0094] [Fig.3] illustrates in its lower part b) an example of application of the different signals to the wires and guides according to the state of the art to travel, in the case of an accelerometer type sensor, the linear trajectories to the two traps and in its upper part a) the position of the two clouds corresponding to chosen times.

[0095] [Fig.4] illustrates a first variant of a matrix chip according to the state of the art allowing the production of several elementary sensors.

[0096] [Fig.4bis] illustrates a second variant of matrix chip according to the state of the art allowing the production of several elementary sensors.

[0097] [Fig.5] illustrates the principle of the sensor according to the invention which comprises a first elementary sensor illustrated in part a) and a second elementary sensor illustrated in part b).

[0098] [Fig.6] illustrates the time course of the measurements carried out with the two sensors, SENA for the upper part a) and SENB for the lower part b).

[0099] [Fig.7] illustrates the ultracold atom interferometric inertial sensor of the accelerometer type according to the invention.

[0100] [Fig.8] illustrates an example of chip topology comprising two crossing ribbons compatible with the production of elementary sensors according to the invention.

[0101] [Fig.9] illustrates another example of chip topology comprising a flared wire compatible with the production of elementary sensors according to the invention.

[0102] [Fig. 10] illustrates an example of implementation for carrying out two acceleration measurements along the same axis, using two elementary sensors on the same axis. chip according to the invention. For clarity, only one pair of guides is shown, the guides parallel to the Y axis, for a measurement of an acceleration along the X axis. DETAILED DESCRIPTION OF THE INVENTION

[0103] The idea of ​​the invention is to use two elementary sensors configured in a particular way to eliminate measurement bias.

[0104] The formula (1) stated above corresponds for example to a SENA sensor for which the trajectory TRIA is traveled by the cloud CL1 and the trajectory TR2A traveled by the cloud CL2, as illustrated [Fig.5] upper part a)). The cloud CL1A leaves the starting point DPA in a starting direction SIA (and the cloud CL2A leaves the starting point in a starting direction S2A not shown in [Fig.5]). In [Fig.5] a) the starting direction SIA is upwards (the direction S2A is downwards). The abscissas of the extreme points of the trajectories TRIA (cloud CL1A in state la>) and TR2A (cloud CL2A in state lb>) of SENA are respectively called xaA and x bA, with Axa = Ix^ - xbAl.

[0105] For an accelerometer, the sign of the phase shift induced by the presence of an acceleration during the interferometry phase depends on the way in which the two clouds of atoms travel the two trajectories. The sign of the phase shift depends on the direction in which the two clouds depart from the starting point.

[0106] Consider two cold atom accelerometers SENA and SENB arranged on the same chip with separation of the traps using a movable trapping potential, each being oriented so that it is sensitive to accelerations along the same axis (e.g. the X axis in the example). This means that the line DrA on which the trajectories TRIA and TR2A of SENA are based and the line DrB on which the trajectories TR1B and TR2B of SENB are based are parallel to each other and to X. The abscissas of the extreme points of the trajectories TR1B (CL1B cloud in the state la>) and TR2B (CL2B cloud in the state lb>) of SENB are respectively called xbB and xaB, with Axb= lxbB - xaBl (see part b) of [Fig.5]).

[0107] Let us consider that for the SENB sensor the departure direction S1B of the first cloud CL1B is opposite to the departure direction SIA of the first cloud CL1A of SENA. We have S1B = -SIA, that is to say that the first cloud CL1B here goes downwards. Of course this condition is also respected for the second clouds CL2A and CL2B. This is illustrated [Fig.5] lower part b). The cloud CL1B leaving the starting point DPB in the direction S1B goes downwards.

[0108] When the directions of travel of the atomic states used on these sensors are reversed as illustrated 5, each accelerometer will accumulate the same phase shift to within a difference in sign.

[0109] This result is obtained for trajectories parallel to each other, of the same length and traveled simultaneously and so that the direction of departure of the clouds CL1A and CL1B are opposite: Axa=Axb=Ax.

[0110] [Fig.6] illustrates the time course of the measurements carried out with the two sensors, SENA for the upper part a) and SENB for the lower part b). After a first ir / 2 pulse (duration r) the two pairs of clouds are separated simultaneously from ti (start of the Ramsey sequence), then are kept separated for a time Ts, then are recombined. Finally a second ir / 2 pulse ends the Ramsay sequence, of total duration TR. By carrying out two measurements simultaneously (one for each sensor), it is possible to eliminate the presence of measurement bias as explained later.

[0111] Let ^tot-SA be the phase accumulated during the interferometric sequence corresponding to figure 5 part a) and ^tot-SB the phase accumulated during the interferometric sequence corresponding to [Fig.5] part b). We have by taking formula (1):

[0112] {p -(p + y = ) I7 (2A) ' tot_SA * dock Tace l\ab ft / JB

[0113] tp -(py = 'early__SB' dock acc l \ ao # / JB

[0114] Considering that the measurement is tainted by a bias, we have:

[0115] [01161 =

[0117] With additional phase bias due to the bias present in any measurement (so-called “additive” bias).

[0118] If a (non-negligible) bias is present, the determination of ^tot-SA - ^tot-SB allows it to be eliminated:

[0119] - m )(3) 'acc 2 \Ytot_SA ^tot—SB

[0120] Thus, to carry out a measurement enabling the bias to be eliminated, two interferometric sequences must be carried out in parallel with two sensors on the same chip.

[0121] The ultracold atom interferometric inertial sensor 10 according to the invention uses this property and is illustrated [Fig.7]. It comprises an atom chip ACh placed in a vacuum chamber, comprising an XY plane called the measurement plane normal to the Z axis. The chip comprises at least one set of a first elementary sensor SENA and a second elementary sensor SENB. Each elementary sensor comprises at least a first pair of waveguides (CPWX1, CPWX2) parallel to each other and at least a second pair of waveguides (CPWY'1, CPWY'2) parallel to each other and intersecting with the first pair. The intersection of the two pairs defines a parallelogram. The atom chip also comprises a group of one or more IEC conductive elements.

[0122] The sensor further comprises an ACG atom generating device configured to generate an initial cloud of ultracold atoms associated with the first elementary sensor SENA and an initial cloud CLB of ultracold atoms associated with the second elementary sensor SENB. The initial clouds are located near said XY plane of the chip. The sensor also comprises a generator GB of a homogeneous magnetic field Bc and a power supply device PSD. The power supply device comprises at least one microwave generator GMW configured to apply microwave signals to the waveguides and at least one direct current generator configured to apply direct currents to the conductive elements. The magnetic field generator GB and the power supply device PSD being configured to apply the magnetic field Bc, the direct currents and the microwave signals according to a predetermined sequence.

[0123] The arrangement of the group of one or more conductive elements and the predetermined sequence are configured, during the implementation of each sensor, to carry out the different steps described above on the clouds.

[0124] This involves firstly generating an initial trapping potential Vini of said initial cloud of ultracold atoms, then spatially separating the initial cloud into a first cloud of ultracold atoms (CL1A for SENA and CL1B for SNEB) in a first internal state from a second cloud of ultracold atoms (CL2A for SENA and CL2B for SENB) in a second internal state, respectively forming a first ultracold atom trap (T1A for SENA and T1B for SENB) and a second ultracold atom trap (T2A for SENA and T2B for SENB). In addition, the first cloud, the second cloud and the starting point are aligned along a measurement axis:

[0125] Starting point DPA for SENA with CL1A, DPA and CL2A aligned along the line DrA.

[0126] Starting point DPB for SENB with CL1B, DPB and CL2B aligned along the line DrB.

[0127] The starting points DPA for SENA and DPB for SENB are preferentially located at the center of the parallelogram, which is a rectangle when the pairs are perpendicular (axis Y'=Y).

[0128] As illustrated [Fig.5] for each sensor the first cloud, the second cloud and the starting point are aligned along a measurement axis.

[0129] Then the separation of the first and second traps is maintained for a time T s, then the traps are recombined at the starting point, so that the first and second clouds respectively travel a first and a second linear trajectory parallel to the XY plane from the starting point.

[0130] The arrangement of the group of one or more conductive elements of each sensor and the sequence of application of direct currents in conductive elements, microwave signals in the guides and the magnetic field to the neighborhood of the atomic chip Ach are further configured so that the trajectories (TRIA and TR2A) associated with the first elementary sensor and the trajectories (TR1B and TR2B) associated with the second elementary sensor are parallel to each other, of the same length and traveled simultaneously. In addition, the departure direction of the first clouds of the first and second elementary sensors respectively are opposite. Let SIA be the departure direction of CL1A and S2A the departure direction of CL1B, we have S1B = -SIA. The same condition is also respected for the second clouds CL2A and CL2B.

[0131] The sensor further comprises a detection system SDET adapted to measure at least a first phase ^io^sa of the first elementary sensor and a second phase Vtot-SB of the second elementary sensor. The acceleration ac is determined from a difference between the first and second phases (property expressed with formula (3)). When implementing the sensor according to the invention, a double simultaneous measurement of the same parameter is therefore carried out, the phase accumulated during the interferometric sequence.

[0132] By using the property of the interferometer explained by formulas (2A) (2B) and (3) the acceleration thus determined is corrected for the so-called “additive” bias which affects each measurement in an identical manner. This reduction in measurement bias leads to a reduced measurement uncertainty of the accelerometer according to the invention, i.e. better sensitivity, which makes it possible to obtain longer inertial navigation times.

[0133] Preferably, the first and second pairs of waveguides are perpendicular to each other. For example, the guides of the first pair are parallel to the X axis and the guides of the second pair are parallel to the Y'=Y axis. The set of two accelerometers (SENA and SENB) is thus capable of measuring accelerations along two orthogonal axes. They measure one or the other of the accelerations without bias depending on the sequence applied.

[0134] According to one embodiment, the sensor comprises at least two assemblies configured respectively to carry out an acceleration measurement along two orthogonal axes. Thus, the measurement of the acceleration acx along X without bias is carried out by the first assembly and the measurement of acy along Y bias is carried out by the second assembly. The measurements of acx and acy without bias can thus be carried out simultaneously.

[0135] The sensor according to the invention is compatible with any atomic chip geometry comprising the two pairs of microwave guides allowing the generation of so-called dressed traps (see state of the art), i.e. the generation of a trapping potential, and the selective separation of these traps as a function of the internal states along a linear trajectory as described previously.

[0136] Thus the IEC conductive elements of the chip can be arranged in different ways. What matters is the generation of a trapping potential resulting from the combination of the DC fields generated by the conductive elements carrying direct currents and the microwave fields generated by the waveguides carrying microwave signals, this trapping potential being able to be displaced to achieve the separation and therefore the linear trajectory.

[0137] For example, the sensor according to the invention is compatible with all chip geometries described in the state of the art, associated with the different addressing modes, i.e. with particular sequences of application of the magnetic field, direct currents and microwave fields.

[0138] Conductive element geometries have been described for gyrometer applications. These geometries are applicable here since they allow separation of the two clouds from a starting point. The accelerometer does not require movement of the traps along a closed trajectory.

[0139] Typically according to the state of the art the initial trapping potential is generated by applying direct currents in the conductive elements.

[0140] According to one embodiment, these conductive elements are two wires which cross at point C, as illustrated [Fig.7] (point CA for SENA and CB for SENB).

[0141] According to another embodiment, the conductive elements comprise two ribbons W1 and W2 which cross as illustrated [Fig.8] and described in document US2022 / 0397397.

[0142] According to another embodiment the conductive elements comprise a flared wire W2 as illustrated [Fig.9] and described in document US2022 / 0397396. The additional pairs of guides (CPWX1', CPWX2') and (CPWY1', CPWY2'), the second wire W1 and the loading wires W1C and W2C illustrated [Fig.9] are optional for the accelerometer according to the invention.

[0143] The separation of the two traps T1 and T2 is carried out by applying microwave signals to the different waveguides.

[0144] According to one embodiment, the separation is maintained by maintaining the microwave signals at the different waveguides.

[0145] According to another embodiment, once the clouds have been separated, the microwave signals applied to the waveguides are replaced by direct currents. This embodiment is described for the gyrometer application in document FR2306475 (not yet published on the filing date). The PSD power supply device is in this case further configured to apply direct currents to the waveguides. This replacement makes it possible to reduce the measurement noise.

[0146] The sensor according to the invention is thus independent of the geometry of the atomic chip and of the method of maintaining the traps used. The chip may have another geometry of conductive elements associated with the waveguides not described in the aforementioned documents, as well as additional waveguides. The implementation of the sensor may use other methods (sequences) of separation / maintenance allowing the realization of a displaceable trapping potential (i.e. whose minimum can be moved) than those described in the aforementioned documents.

[0147] In practice, an example of implementation of the realization of two acceleration measurements along the same axis, using two elementary sensors on the same chip by moving each state antisymmetrically is illustrated [Fig. 10], which is based on the operating principle of an accelerometer illustrated [Fig.3]. For clarity, only one pair of guides is shown, the guides parallel to the Y axis, for a measurement of an acceleration along the X axis.

[0148] The Vini potential is generated for each sensor by two wires, (WIA and W2A) for SENA and (WIB and W2B) for SENB, and carried by direct currents. The two wires cross at a point placed at an equal distance from two guides. Thus the DC wires of the same sensor (with an external bias field Bc) make it possible to create a magnetic trap to manipulate and conserve the atoms, in particular during the interferometry phase.

[0149] Coplanar waveguides allow, when a microwave signal is injected inside (at the right power and frequency), to selectively move the atoms according to the electronic state in which it is found by pushing them.

[0150] In [Fig. 10], different instants of the interferometry sequence are schematically represented to better visualize the displacements induced on the atomic states due to the different DC and microwave currents to be injected at the same time on the two sensors.

[0151] At t = 0 the atoms of the SENA sensor are in the state la> and those of the SENB sensor also. The cloud of each of the sensors was prepared in the same conditions so that they have the same properties (number of atoms, temperature, state ...). Each sensor has a cloud of atoms trapped inside a magnetic trap created by the two DC wires of the sensor which corresponds to it. The coplanar waveguides are currently switched off. The interferometry sequence of each sensor begins simultaneously and in the same conditions at t = 0; by a pulse n / 2 of duration T (for each sensor) and which ends at tb as illustrated [Fig.6]. This pulse will put each of the clouds in a coherent superposition with equal weights of the two states of the interferometer.

[0152] Between ti (part a) of [Fig. 10]) and t2 (part b) of [Fig. 10]), we gradually turn on the coplanar waveguides of the two sensors as shown in [Fig. 10]. Indeed, we need to selectively move each state in opposite directions. The subtlety lies in the frequencies used in the waveguides of each sensor:

[0153] In the coplanar waveguides of the SENA sensor (on the left), the microwave frequency coa corresponding to the covering of the state la> in the waveguide which is in the positive x, and the microwave frequency cob corresponding to the covering of the state lb> in the waveguide which is in the negative x are injected.

[0154] In the coplanar waveguides of the SENB sensor (on the right), the microwave frequency coa corresponding to the dressing of the state la> in the waveguide which is in the negative x, and the microwave frequency cob corresponding to the dressing of the state lb> in the waveguide which is in the positive x are injected.

[0155] Thus, the SENA sensor will therefore move the state lb> towards the positive x and the state la> towards the negative x while the SENB sensor will move the state la> towards the positive x and the state lb> towards the negative x. The selective movement of the states is therefore done in an antisymmetric manner between the two sensors, that is to say that the starting directions of each type of cloud are opposite for the two sensors.

[0156] After a separation maintenance period, at time t3 (part c) of figure 10 the recombination of the clouds on the two sensors has just ended with the simultaneous extinction of all the waveguides, and the second pulse n / 2 can begin (one for each sensor).

[0157] Finally, we cut all the magnetic fields to free the atoms and we detect the populations of atoms of each sensor to determine ^tot_sA and VtotjSB, then to calculate the acceleration according to X acx from ( ^totjsA - ) by eliminating the bais of measurement.

[0158] Using the pair of guides along X (not shown [Fig. 10]) according to the same reasoning we measure the acceleration acy along the Y axis.

[0159] According to one embodiment, the atomic chip has a matrix structure whose pixels define potential elementary sensors, a set of two sensors (SENA and SENB) comprising two pixels of the matrix. An example of matrix chip topology is illustrated [Fig.4], another [Fig.4bis].

[0160] For the implementation of the invention, two pixels of the matrix are configured to form an assembly according to the invention for the measurement of an acceleration.

[0161] As different microwave signals are applied to the two sensors SENA and SENB, they cannot have the waveguides in common. According to one embodiment, sets of a first and a second elementary sensor have at least one conductive element in common. A topology according to [Fig.4bis] is preferred in which SENA and SENB are chosen in the matrix in order to share the same conductive wire but not the same waveguides.

[0162] According to another aspect, the invention relates to an inertial unit comprising at least two accelerometer-type sensors according to the invention configured to carry out an acceleration measurement respectively along two orthogonal axes, the chip matrix atomic further comprising pixels configured to perform at least one clock measurement and pixels configured to perform rotation speed measurements along three orthogonal axes.

[0163] According to another aspect, the invention relates to a method for measuring an acceleration by an interferometric ultracold atom inertial sensor 10 according to the invention as described previously.

[0164] The method comprises, for each sensor, steps A to F. In a step A, an initial cloud of ultracold atoms associated with the first elementary sensor SENA and an initial cloud CLB of ultracold atoms associated with the second elementary sensor SENB are generated, the initial clouds being located near said XY plane of the chip. In a step B, a homogeneous magnetic field Bc is generated in the vicinity of the chip. Steps A and B can be simultaneous. In a step C, an initial trapping potential Vini of the initial cloud of ultracold atoms is generated (for each sensor). Then, for each sensor, in a step D, a first internal state and a second internal state are initialized by a first pulse tt / 2.

[0165] In a step E for each sensor, the initial cloud is spatially separated into a first cloud (CL1A for SENA, CL1B for SENB) of ultracold atoms in the first internal state from a second cloud (CL2A for SNA, CL2B for SNB) of ultracold atoms in the second internal state, respectively forming a first ultracold atom trap (T1A for SENA, T1B for SENB) and a second ultracold atom trap (T2A for SNA, T2B for SNB). For each sensor, the first cloud, the second cloud and the starting point are aligned along a measurement axis.

[0166] For each sensor, the separation of the first and second traps is maintained for a time Ts and then the traps are recombined at the starting point, so that the first and second clouds respectively travel a first linear trajectory (TRIA for SENA, TR1B for SENB) and a second linear trajectory (TR2A for SNA, TR2B for SNB) from the starting point. The trajectories are parallel to the XY plane of the chip.

[0167] Steps B to F are carried out by applying, according to a predetermined sequence, a homogeneous magnetic field in the vicinity of the chip, direct currents to CEI conductive elements and SMW microwave signals to the waveguides.

[0168] In order to be able to measure the same phase with an opposite sign, the arrangement of the group of elements of one or more conductive elements and the predetermined sequence are further configured so that the trajectories (TRIA and TR2A) associated with the first elementary sensor and the trajectories (TR1B and TR2B) associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously, and so that the starting direction (i.e. at the start from points DPA and DPB) of the first clouds CL1A and CL1B respectively of the first SENA sensor and the second SENB sensor are opposite.

[0169] Then in a step G the first and second internal states are recombined by applying to said ultracold atoms a second pulse ji / 2. In a step H at least a first phase ^tot_sa of the first elementary sensor and a second phase ^tot_SB of the second elementary sensor are measured, the acceleration being determined from the difference between the first and second phases.

[0170] According to one embodiment, step C comprises the application of direct currents in at least one conductive element.

[0171] According to one embodiment, step E and step F comprise the application of microwave signals in a pair of waveguides in order to measure an acceleration along a measurement axis perpendicular to the waveguides of the pair in question.

[0172] According to one embodiment in step E and step F the application of microwave signals comprises:

[0173] -for the first elementary sensor SENA the application of a signal of frequency coa in one of the waveguides and the application of a frequency cob in the other waveguide,

[0174] -for the second SENB sensor the application of a signal of frequency cob in one of the waveguides and the application of a frequency coa in the other waveguide,

[0175] The frequencies coa and cob are determined from the resonance frequencies of the first and second internal states.

[0176] For the SENB sensor, the choice of which frequency is to be applied to which guide is determined to obtain the inversion of the starting directions with respect to those of SNA.

[0177] According to one embodiment, to reduce the measurement noise, step F comprises a sub-step of cutting off microwave signals and a sub-step of applying direct currents to the waveguides, in place of the microwave signals.

Claims

Claims

1. Ultra-cold atony inertial sensor (10) interferometric accelerometer type comprising: • an atomic chip (ACh) placed in a vacuum chamber, comprising an XY plane called the measurement plane normal to a Z axis, and comprising at least one set of a first and a second elementary sensor (SENA, SENB), each elementary sensor comprising: • at least a first pair of waveguides (CPWX1, CPWX2) parallel to each other and at least a second pair of waveguides (CPWY' 1, CPWY'2) parallel to each other and intersecting with the first pair, • a group of one or more conductive elements, • an atom generation device (ACG) configured to generate an initial cloud of ultracold atoms associated with the first elementary sensor (SENA) and an initial cloud (CLB) of ultracold atoms associated with the second elementary sensor (SENB), said initial clouds being located near said XY plane of said chip, • a generator (GB) of a homogeneous magnetic field (Bc), • a power supply device (PSD) comprising at least one microwave generator (GMW) and at least one direct current generator (GDC), the power supply device being configured to apply microwave signals to said waveguides and direct currents to said conductive elements, • the magnetic field generator and the power supply device being configured to apply the magnetic field, the direct currents and the microwave signals in a predetermined sequence, • an arrangement of said group of one or more conductive elements and said sequence being configured, during the implementation of each sensor, to: i. Generate an initial potential for trapping said cloud initial of ultracold atoms defining a starting point, ii. Spatially separating the initial cloud into a first cloud (CL1A, CL1B) of ultracold atoms in a first internal state from a second cloud (CL2A, CL2B) of ultracold atoms in a second internal state, respectively forming a first (T1A, T1B) and second (T2A, T2B) ultracold atom trap, and such that the first cloud, the second cloud and the starting point are aligned along a measurement axis, iii. Maintaining the separation of said first and second traps for a time Ts and then recombining said traps at said starting point, so that said first and second clouds respectively travel a first and a second linear trajectory parallel to the XY plane from the starting point, the arrangement of said group of one or more conductive elements of each sensor and said sequence being further configured so that the trajectories (TRIA, TR2A) associated with the first elementary sensor and the trajectories (TR1B, TR2B) associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously and so that the direction of departure of the first clouds, respectively of the first and second elementary sensors, are opposite, the sensor further comprising a detection system (SDET) adapted to measure at least a first phase (of the first elementary sensor and a second phase (Ptot-Sü) of the second elementary sensor, an acceleration being determined from a difference between the first and second phase.

2. Sensor according to the preceding claim wherein the power supply device PSD is further configured to apply direct currents to said waveguides.

3. Sensor according to one of the preceding claims in which the first and second pair of waveguides are perpendicular to each other, and in which the sensor comprises at least two assemblies configured to carry out an acceleration measurement respectively along two orthogonal axes.

4. Sensor according to one of the preceding claims in which the atomic chip has a matrix structure whose pixels define potential elementary sensors, a set comprising two pixels of the matrix.

5. Sensor according to the preceding claim in which sets of a first and a second elementary sensor have at least one conductive element in common.

6. Inertial unit comprising at least two accelerometer-type sensors according to one of claims 3 to 5 configured to carry out an acceleration measurement respectively along two orthogonal axes, the atomic matrix chip further comprising pixels configured to carry out at least one clock measurement and pixels configured to carry out rotation speed measurements along three orthogonal axes.

7. Method for measuring an acceleration by an interferometric ultracold atom inertial sensor (10) comprising an atom chip (ACh) placed in a vacuum chamber, comprising an XY plane called the measurement plane normal to a Z axis, and comprising at least one set of a first and a second elementary sensor (SENA, SENB), each elementary sensor comprising: • at least one first pair of waveguides (CPWX1, CPWX2) parallel to each other and at least one second pair of waveguides (CPWY' 1, CPWY'2) parallel to each other and intersecting with the first pair, • a group of one or more conductive elements, the method comprising the steps of: A. Generating an initial cloud of ultracold atoms associated with the first elementary sensor (SENA) and an initial cloud (CLB) of ultracold atoms associated with the second elementary sensor (SENB), said initial clouds being located near said XY plane of said chip, B. Generate a homogeneous magnetic field, C. For each sensor generate an initial trapping potential (Vini) of said initial cloud of ultracold atoms, D. For each sensor initialize a first internal state and a second internal state by a first pulse rr / 2, E. For each sensor, spatially separate the initial cloud into a first cloud (CL1A, CL1B) of ultracold atoms in the first internal state from a second cloud (CL2A, CL2B) of ultracold atoms in the second internal state, respectively forming a first (T1A, T1B) and second (T2A, T2B) ultracold atom trap, and such that the first cloud, the second cloud and the starting point are aligned along a measurement axis, F. For each sensor, maintaining the separation of said first and second traps for a time Ts and then recombining said traps at said starting point, so that said first and second clouds respectively travel a first and a second linear trajectory parallel to XY from the starting point, steps B to F being carried out by applying, according to a predetermined sequence, a homogeneous magnetic field, direct currents to said conductive elements and microwave signals (IMw) to said waveguides, the arrangement of said group of elements of one or more conductive elements and said predetermined sequence being further configured so that the trajectories (TRIA, TR2A) associated with the first elementary sensor and the trajectories (TR1B, TR2B) associated with the second elementary sensor are parallel to each other, of the same length, traveled simultaneously, and so that the starting direction of the first clouds,respectively of the first and second sensors, are opposite, G. Recombine said first and second internal states by applying to said ultracold atoms a second pulse 2, H. Measure at least a first phase (Vmja) of the first elementary sensor and a second phase of the second elementary sensor, an acceleration being determined from a difference between the first and second phases.

8. Method according to the preceding claim in which step C comprises the application of direct currents in at least one conductive element.

9. A method according to either of claims 7 or 8 wherein step E and step F comprise applying microwave signals to a pair of waveguides to measure acceleration along a measurement axis perpendicular to the waveguides of said pair.

10. Method according to the preceding claim in which in step E and step F the application of microwave signals comprises: • for the first elementary sensor the application of a signal of frequency coa in one of the waveguides and the application of a frequency cob in the other waveguide, • for the second sensor the application of a signal of frequency cob in one of the waveguides and the application of a frequency coa in the other waveguide, said frequencies coa and cob being determined from the resonance frequencies of said first and second internal states.

11. Method according to one of claims 6 to 10 in which step F comprises a sub-step of cutting off the microwave signals and a sub-step of applying direct currents to said waveguides.

Images