Method for correcting attitude provided by dead reckoning navigation system by means of relative positioning system

Abstract

The invention relates to an attitude correction method (107), comprising estimating (116) a plurality of estimated positions of a dead reckoning navigation system (22) in any fixed reference frame at the time of determination by means of the dead reckoning navigation system (22), obtaining (114) the position of the dead reckoning navigation system (22) in a predetermined fixed reference frame, evaluated by a relative positioning system (24), at the time of determination, and inferring (124) at least one attitude correction parameter by minimizing a cost function comparing the evaluated position with the estimated position corrected by means of the correction parameter or each correction parameter.

Description

Technical Field

[0001] The present invention relates to dead reckoning navigation technology, and more particularly to a technique for resetting the attitude provided by a dead reckoning navigation system. It has advantageous applications in locomotion in urban or indoor environments (i.e., within buildings). Background Art

[0002] Nowadays, it is common to track the position of an object by multilateral methods, by measuring the distance between a receiver attached to the object and at least three sources whose positions in the environment are known. This is the case, for example, with positioning systems of the GNSS (Global Navigation Satellite System, such as GPS) type or those that benefit from an infrastructure of wireless communication networks (such as Wi-Fi networks, GSM networks, etc.). However, these methods are very limited since they do not ensure the availability and accuracy of the information, both of which are affected by the shielding that may exist between the source and the receiver. Therefore, their use in urban or indoor environments requires the deployment of an expensive infrastructure with many sources distributed throughout the environment. They also rely on external technologies such as satellites of GNSS, which may not be available or even deliberately interfered with.

[0003] Alternatively, methods known as dead reckoning navigation methods are also known, which allow tracking the relative position of an object in any environment, thanks to motion sensors measuring the object's motion. The relative position is the position of an object in space relative to a given point and coordinate system at the time of its initialization. In addition to the position, these methods can also obtain the orientation (also called "pose") of the object relative to the same initial coordinate system, which is represented in three dimensions by the Euler angles (roll Pitch θ, yaw ψ) is given by , and in dimension 2, by heading ψ. These methods are preferred when moving in environments where position tracking by multilateral methods is difficult, such as urban or indoor environments.

[0004] There are different types of dead reckoning navigation. The most common is the one known as "simple" inertial navigation, used in heavy applications such as fighter or passenger aircraft, submarines, ships, etc. It is based on an inertial unit, usually comprising at least three accelerometers and three gyroscopes arranged on three axes. Usually, the gyroscopes "maintain" a coordinate system in which the double time integration of the accelerometer measurements makes it possible to estimate the movement. It is well known that in order to use this "simple" inertial navigation method, very high-precision sensors are required. In fact, the double time integration of the acceleration measurements means that a constant acceleration error generates a position error that increases proportionally to the square of the time.

[0005] Another known dead reckoning navigation technique is one where an external source (e.g., a car's odometer, a boat's log, or an airplane's pitot tube) provides velocity vector information in the object's coordinate system. This velocity vector information is then simply integrated and combined with the object's orientation information, particularly its heading, to know its trajectory. Time drift is less sensitive to the same measurement inaccuracies of the sensors.

[0006] Typically, the initial attitude is known, for example by an initial "alignment" of the inertial device or by powering of sensors other than inertial sensors (e.g. magnetic sensors). However, measurement inaccuracies of inertial sensors can lead to a temporal drift in the measured attitude, which makes the knowledge of the initial attitude obsolete after a more or less long time, depending on the accuracy of the sensors used, and leads to inaccuracies in the object's position. For example, a 1% error in the heading measurement after moving 100 meters can lead to an object position inaccurate by 1 meter.

[0007] To solve this problem, it is known to periodically reset the measurements of a dead reckoning navigation system with the aid of another positioning system in order to best maintain the orientation information, in particular the heading information, of the object. For example, it is known from WO 2019 / 020961 to reset the orientation information by using magnetic heading measurements obtained by a magnetometer carried by the object.

[0008] However, this solution is not entirely satisfactory. In fact, magnetometers have their own inaccuracies and drifts, which means that their measurements are sometimes not reliable enough to be used as a basis for reset. In addition, this solution requires the magnetometer to be integrated into the object whose position is to be tracked, which increases its cost. Summary of the invention

[0009] One object of the invention is to propose a simple and economical solution for resetting the attitude provided by a dead reckoning navigation system. Another object is to allow such resetting to be carried out with great accuracy over short distances. Another object is to accurately track people moving inside a building in a simple and economical way.

[0010] To this end, according to a first aspect, the present invention relates to an attitude resetting method for resetting an attitude provided by a dead reckoning navigation system, the resetting method comprising the following steps:

[0011] - estimating, by the dead reckoning navigation system, for each of a plurality of determined times included in a movement interval in which the dead reckoning navigation system moves along the movement trajectory, an estimated position of the dead reckoning navigation system in an arbitrary fixed coordinate system at the determined time,

[0012] - obtaining, for each determination instant, an estimated position of the dead reckoning navigation system in a predetermined fixed coordinate system at said determination instant, said estimated position having been estimated by the relative positioning system, and

[0013] - inferring at least one pose reset parameter by minimizing a cost function comparing the evaluated position with an estimated position corrected by means of the or each reset parameter.

[0014] According to a specific embodiment of the present invention, the posture resetting method further has one or more of the following features, which can be used alone or in any technically possible combination:

[0015] - a dead reckoning navigation system is worn by the pedestrian, the dead reckoning navigation system preferably being attached to the pedestrian's foot or ankle;

[0016] - the dead reckoning navigation system comprises a motion sensor for measuring the motion of the dead reckoning navigation system and a processing unit for inferring the attitude and position of the dead reckoning navigation system from the measured motion;

[0017] - a relative positioning system selected from the group consisting of a multi-angle system, a multilateral system, a map matching system and a vision positioning system, the relative positioning system preferably comprising an ultra-wideband telemetry device;

[0018] The arbitrary fixed coordinate system and the predetermined fixed coordinate system share a common axis, and the attitude resetting parameters consist of parameters (preferably angles) that modify the orientation by rotating around said common axis;

[0019] - the common axis is the vertical axis;

[0020] - a cost function representing the average geometric deviation between the evaluation position and the estimated position after applying a geometric transformation comprising a rotation and preferably a translation to the evaluation position or the estimated position;

[0021] - rotation about an axis of rotation and translation in a direction orthogonal to said axis of rotation;

[0022] - the step of inferring a reset parameter comprises: calculating candidate values ​​of the reset parameter, evaluating a precision value of the candidate value, comparing the precision value with a previous precision value associated with a previous reset parameter, and determining the reset parameter based on the comparison result, the reset parameter depending on the candidate value and the previous reset parameter.

[0023] - The accuracy value is a function of the uncertainty in the estimated and evaluated positions and / or the average geometric deviation between the evaluated and estimated positions after applying the candidate values ​​of the reset parameters.

[0024] - Inferring reset parameters includes the following sub-steps:

[0025] oa) calculating a first candidate value of a reset parameter by minimizing a cost function comparing the evaluated position at N determined moments with the estimated position at said N determined moments corrected by means of the reset parameter,

[0026] ob) calculating a first precision value associated with said first candidate value,

[0027] oc) calculating a second candidate value of the reset parameter by minimizing a cost function comparing the evaluated positions at N-1 determination moments corresponding to the N determination moments minus the oldest determination moment with the estimated positions at the N-1 determination moments corrected by means of the reset parameter or each reset parameter,

[0028] od) calculating a second precision value associated with the second candidate value,

[0029] oe) compares the first precision value and the second precision value, and

[0030] of) when the first precision value reflects the best precision, selecting the first candidate value;

[0031] When the precision value reflecting the best precision consists of the second precision value, inferring the reset parameter comprises the following additional sub-steps:

[0032] o delete the oldest determined moment's evaluated and estimated positions, and

[0033] o repeating sub-steps a) to e), wherein the number N is decreased by 1; and

[0034] -Dead reckoning navigation systems have an accuracy of a few percent in terms of distance traveled, e.g. between 1% and 3%

[0035] The accuracy in heading drift is about tens of degrees per hour, for example, between 30 and 80 degrees per hour, and the position accuracy of the relative positioning system is tens of centimeters, for example, between 20 cm and 1 m.

[0036] According to a second aspect, the present invention also relates to a method for locating an object in a predefined space, the method comprising the following steps:

[0037] - Activate the dead reckoning navigation system,

[0038] - resetting the attitude of a dead reckoning navigation system using a relative positioning system comprising an infrastructure installed at an access point in a predefined space to obtain attitude resetting parameters, said attitude resetting implementing the attitude resetting method according to any of the preceding claims,

[0039] - resetting the position of the dead reckoning navigation system using a relative positioning system comprising infrastructure installed at said access points of the predefined space to obtain position resetting parameters, and

[0040] - calculating, by the dead reckoning navigation system, a calculated position of the dead reckoning navigation system in the predetermined coordinate system with the aid of the attitude reset parameter and the position reset parameter.

[0041] According to a specific embodiment of the present invention, the positioning method also has the following characteristics:

[0042] The infrastructure is loaded onto a vehicle which is itself equipped with a positioning and orientation system making it possible to calculate the position of the infrastructure in a predetermined coordinate system.

[0043] According to a third aspect, the present invention also relates to a dead reckoning navigation system, comprising a motion sensor for measuring the motion of the dead reckoning navigation system and a processing unit for inferring the attitude and position of the dead reckoning navigation system in a fixed coordinate system from the measured motion, the processing unit being configured to implement the attitude resetting method according to the first aspect to reset the attitude.

[0044] According to a fourth aspect, the present invention relates to a computer program product comprising code instructions for executing the posture resetting method according to the first aspect when said program is executed by a processor.

[0045] Finally, according to a fifth aspect, the present invention relates to a storage device readable by a computer device, on which a computer program product including code instructions for executing the posture resetting method according to the first aspect is recorded. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Other features and advantages of the invention will emerge on reading the following description given by way of example only and with reference to the accompanying drawings, in which:

[0047] - Figure 1 is a top view of a system for locating an object in a predefined space according to an exemplary embodiment of the present invention,

[0048] - Figure 2 yes Figure 1 A perspective view of a detail of the positioning system,

[0049] - Figure 3 yes Figure 1 Schematic diagram of the positioning box of the positioning system,

[0050] - Figure 4 It is shown by Figure 1 A schematic diagram of an example of a method for locating an object in a predefined space implemented by a system of

[0051] - Figure 5 It is shown Figure 4 A schematic diagram of the heading resetting step of the method, and

[0052] - Figure 6 It shows the inference Figure 5 Schematic diagram of the sub-steps of the heading reset parameters of the heading reset step. DETAILED DESCRIPTION

[0053] Figure 1 The positioning system 10 represented in FIG. 1 is intended to locate an object of interest 12 within a predefined space 14. To this end, the positioning system 10 comprises a plurality of positioning boxes 16, each of which is attached to a respective object of interest 12. Here, the positioning system 10 also comprises a positioning infrastructure 18 arranged at least at one access point 19 of the predefined space 14.

[0054] refer to Figure 2 , each object of interest 12 here consists of a pedestrian. The invention is particularly advantageous in this use, since it allows the volume of the box 16 to be greatly reduced, so that the pedestrian can easily wear the box 16 ergonomically. As a variant (not represented), the object of interest 12 consists of any mobile object for which knowledge of the position is required, such as a wheeled vehicle, a drone, etc.

[0055] The predefined space 14 is usually constituted by the interior of a building. For example, if it is an industrial site, then the pedestrian 12 is usually a technician working at the industrial site. As a variant, if the predefined space 14 is an intervention site, such as a burning building or a hostage situation, then the pedestrian 12 is a firefighter or infantryman intervening in the site.

[0056] Each positioning box 16 is usually attached to a limb of the pedestrian 12, here a leg, preferably a foot or an ankle. Figure 3 As shown, each positioning box 16 comprises an attachment member 20, here constituted by a bracelet, for example with a hook and loop strip that goes around a limb and allows a secure connection. As a variant (not shown), the attachment member 20 is constituted by any element that allows the connection with the object 12 of interest to be fixed to the positioning box 16.

[0057] Still reference Figure 3 , the positioning box 12 includes a dead reckoning navigation system 22. It also includes a relative positioning system 24. In the example shown, it also includes a communication system 26, typically a wireless communication system, for communicating between the positioning box 12 and an external device (such as a mobile terminal 29 ( Figure 2 ), such as a multifunctional mobile device, or even a remote server (not shown). Optionally, it includes a storage module 28.

[0058] The dead reckoning navigation system 22 includes a motion sensor 30 for measuring the motion of the dead reckoning navigation system 22 and a processing unit 32 for inferring the attitude and position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system (e.g., the East-North-Up (ENU) coordinate system) from the measured motion. "Fixed coordinate system" here and hereinafter means that the coordinate system is stationary in the earth's reference system.

[0059] In the example described here, the dead reckoning navigation system 22 is configured to operate in a two-dimensional coordinate system and only provides the heading and two position coordinates of the dead reckoning navigation system 22 in the horizontal plane of the predetermined coordinate system. As a variant (not shown), the dead reckoning navigation system 22 is configured to operate in a three-dimensional coordinate system and thus provides the roll angle, pitch angle and yaw angle of the dead reckoning navigation system 22, as well as its three position coordinates in the predetermined coordinate system. Those skilled in the art will be able to easily convert the examples given here for the two-dimensional case to the three-dimensional case.

[0060] The motion sensor 30 comprises a gyroscope 40 for measuring the angular velocity of the dead reckoning navigation system 22 according to a system with three orthogonal axes defining a mobile coordinate system fixed to the box 16, i.e. measuring the three components of the angular velocity vector in said mobile coordinate system. It is therefore understood that the gyroscope 40 may actually designate a set of three gyroscopes associated with one of the three axes, in particular in three axes (i.e. each gyroscope is capable of measuring one of the three components of the angular velocity vector).

[0061] The motion sensor 30 also comprises means 42 for acquiring the linear velocity, ie the linear velocity of the motion, of the dead reckoning navigation system 22. This acquisition means 42 makes it possible to obtain the linear velocity directly or indirectly and can therefore be of various types.

[0062] For example, the acquisition means 42 may consist of one or more accelerometers (not shown). These accelerometers are advantageously arranged in three axes, usually according to the same system with three orthogonal axes as the gyroscope 40. They are sensitive to external forces other than gravity applied to the sensor 30 and can measure a specific acceleration. The linear velocity is then obtained by integrating this acceleration over time.

[0063] As a variant, when the object 12 is a wheeled vehicle, the acquisition means 42 are composed of at least two odometers, one for each wheel of the vehicle, for example the two rear wheels. An odometer is a device capable of measuring the speed of a wheel by counting the number of revolutions ("tachometer"). Usually, the odometer has a part fixed to the wheel (for example a magnet) and detects each passage of this fixed part (called "top") so as to count the number of revolutions per unit time, that is, the frequency of rotation. Other techniques are known, such as the optical detection of the imprint on the wheel, or the magnetometer of patent FR 2 939 514, which detects the rotation of a metal object (such as a wheel). Here, the "speed" of the wheel is a scalar, that is, the norm of the speed of the wheel in the Earth's reference system (assuming no slip). If the spokes of the wheel are known, this speed norm can be estimated by measuring the frequency of rotation.

[0064] Optionally, the motion sensor 30 also includes a stride detector (not shown) to detect when the feet of the pedestrian 12 are on the ground, for example as described in document WO 2017 / 060660.

[0065] In the example shown, the processing unit 32 is constituted by a programmable machine, such as a DSP (digital signal processor) or a microcontroller. It comprises a processor or CPU (central processing unit) 44 and a memory 46 of the RAM (random access memory) and / or ROM (read-only memory) type. The processor 44 is configured to execute instructions loaded into the memory 46. When the dead reckoning navigation system 22 is powered on, the processor 44 is able to read the instructions from the memory 46 and execute them. These instructions form a computer program that causes the processor 44 to calculate the heading and position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system, for example by implementing the method described in WO 2017 / 060660 or FR 2 939 514.

[0066] As a variant (not represented), the processing unit 32 is constituted by a dedicated machine or component, such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit).

[0067] The processing unit 32 also includes a buffer memory 49 for temporarily storing information required to calculate the heading position and position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system.

[0068] Optionally, the dead reckoning navigation system 22 also includes a network of magnetometers 48 linked to the box 16, i.e., each of them has substantially the same motion as the box 16 in the earth reference system and is spatially spaced from each other. Each magnetometer 48 is a three-axis magnetometer capable of measuring magnetic fields along three axes. To this end, each magnetometer 48 is typically composed of three single-axis magnetometers (not shown) oriented along axes that are substantially perpendicular to each other. These axes are preferably the same as the axes of a system having three orthogonal axes of the gyroscope 40.

[0069] Due to its specific geometry, the network of magnetometers 48 enables determination of the spatial gradient of the measured magnetic field, in particular the coefficient of this gradient along each axis of the mobile coordinate system linked to the box 16, at each measurement moment of the magnetometers 48. Each coefficient of said gradient is determined, for example, by a method using vector measurements of the magnetic field made by the magnetometers 48, associated with an optimization method of the least squares or median filter type, or associated with the intrinsic properties of the magnetic field described by the Maxwell equations. However, any other conventional method suitable for calculating the coefficient of the spatial gradient of the magnetic field is suitable.

[0070] The processing unit 32 is then typically configured to adjust the determination of the velocity of the dead reckoning navigation system 22 by implementing the method described in EP 2 541 199 .

[0071] The accuracy of the dead reckoning navigation system 22 in terms of distance traveled is typically on the order of a few percent, such as between 1% and 3%, and the accuracy in terms of heading drift is typically on the order of tens of degrees per hour, such as between 30 degrees and 80 degrees per hour.

[0072] The relative positioning system 24 is capable of identifying the relative position of a point fixed to the dead reckoning navigation system 22 relative to a reference system having a known position in a predetermined fixed coordinate system, and is capable of inferring therefrom the position of the dead reckoning navigation system 22 in the predetermined fixed coordinate system. To this end, the relative positioning system 24 includes a sensor 50 capable of measuring a parameter that enables the positioning of the safety point relative to the reference system, and a processing unit 52 for inferring the position of the dead reckoning navigation system 22 in the predetermined fixed coordinate system based on the parameter.

[0073] Preferably, the relative positioning system 24 is constituted by a multilateral system, in particular by a neighboring multilateral system. The infrastructure 18 then comprises at least three beacons 54 ( Figure 2 ) (or "anchor"), each beacon has a known position in a predetermined fixed coordinate system, and the relative positioning system 24 is configured to communicate with it. To this end, the sensor 50 is typically composed of a wireless communication system capable of communicating with the beacons 54 and inferring the distance to each beacon 54, for example by measuring the two-way ranging (TWR) distance.

[0074] Preferably, the sensor 50 is constituted by an ultra-wideband telemetry device capable of communicating with the beacon 54 and measuring its distance to the beacon 54 via an ultra-wideband (UWB) protocol, which allows a measurement accuracy of a position of the order of 10 centimeters, typically between 20 cm and 1 m. As known to those skilled in the art, the ultra-wideband protocol is a radio communication protocol based on the transmission of extremely short pulses (nanoseconds) over a wide spectrum. Therefore, depending on the channel used, the communication benefits from a wide bandwidth (500 MHz to 1350 MHz) in the range of 0.5 GHz to 9.5 GHz (center frequency). As a variant, the sensor 50 can communicate with the beacon 54 via a Bluetooth protocol or a Wi-Fi protocol.

[0075] As a variant, the multilateral system consists of a GNSS system.

[0076] The processing unit 52 is then configured to infer the relative position of the sensor 50 in the predetermined fixed coordinate system from the range measurements and the known position of the beacon 54, typically by multilateral methods or by optimization, and to infer the position of the dead reckoning navigation system 22 in the predetermined fixed coordinate system from that position and from the position of the sensor 50 relative to the dead reckoning navigation system 22.

[0077] According to another embodiment (not shown), the relative positioning system 24 is constituted by a multi-angle system, the sensor 50 being able to measure at least one angle between the viewing directions of two beacons, each beacon having a known position in a predetermined fixed coordinate system. The processing unit 52 is then configured to infer the position of the sensor 50 in the predetermined fixed coordinate system from one or more angle measurements and from the known positions of the beacons 54, typically from multiple angles, and to infer the position of the dead reckoning navigation system 22 in the predetermined fixed coordinate system from this position and from the position of the sensor 50 relative to the dead reckoning navigation system 22.

[0078] According to yet another embodiment (not represented), the relative positioning system 24 consists of a map matching system. The sensor 50 is then able to measure parameters of the environment, such as the topography or the magnetic field, and the processing unit 52 is able to match the measurement results with a map of said parameters stored in its memory in order to deduce therefrom the position of the sensor 50 in a predetermined fixed coordinate system.

[0079] According to a fourth embodiment (not represented), the relative positioning system 24 is constituted by a visual positioning system. The sensor 50 is then constituted by an imager associated with an image processing system. The imager is configured to acquire images of the environment and the processing system to detect in each image salient points, such as lines of sight, whose positions are known in a predetermined fixed coordinate system. The processing unit 52 is then configured to infer therefrom the relative position of the sensor relative to the salient points and from this relative position and from the known positions of the salient points and the position of the sensor 50 relative to the dead reckoning navigation system 22 the position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system.

[0080] As a variant (not shown), the imager is fixed (it belongs to the infrastructure 18) and it is a prominent point (usually the line of sight) carried by the box 16. As another variant, a motion capture configuration can be provided in which multiple targets are carried by the box 16 and detected by the fixed imager.

[0081] In the example shown, the processing unit 52 is constituted by a programmable machine, such as a DSP (digital signal processor) or a microcontroller. It includes a processor or CPU (central processing unit) 56 and a memory 58 of the RAM (random access memory) and / or ROM (read-only memory) type. The processor 56 is configured to execute instructions loaded into the memory 58. When the relative positioning system 24 is powered on, the processor 56 is able to read the instructions from the memory 58 and execute them. These instructions form a computer program that enables the processor 56 to calculate the position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system based on the parameters measured by the sensor 50.

[0082] As a variant (not represented), the processing unit 52 is constituted by a machine or a dedicated component, such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit).

[0083] In the example shown, the processing unit 52 of the relative positioning system 24 is distinct from the processing unit 32 of the dead reckoning navigation system 22. As a variant (not shown), these processing units 32, 52 are merged.

[0084] The communication system 26 is configured to implement short-range wireless communication, such as Bluetooth or Wi-Fi (particularly in an embodiment with a mobile terminal 29) and / or to connect to a mobile network (typically UMTS / LTE / 5G) for long-range communication. As a variant (not shown), the communication system 26 is, for example, a wired connector technology (typically USB) for transferring data from the storage module 28 to another storage module, typically a mobile terminal 29.

[0085] For example, the communication system 26 is configured to send the position calculated by the dead reckoning navigation system 22 to the mobile terminal 29 so that the mobile terminal 29 displays the position in the interface of the navigation software.

[0086] In the above example, the processing units 32, 52 of the dead reckoning navigation system 22 and of the relative positioning system 24 are integrated into the box 16. As a variant (not represented), at least a part of these processing units 32, 52 is moved, for example in the mobile terminal 29, the infrastructure 18 and / or a remote server (not represented). In other words, at least a part of the steps of calculating the position of the dead reckoning navigation system 22 by the dead reckoning navigation system 22 or by the relative positioning system 24 is performed by the mobile terminal 29, the infrastructure 18 and / or the remote server. The communication system 26 is then configured to send data from the motion sensor 30 and / or the sensor 50 to the mobile terminal 29, the infrastructure 18 and / or the remote server. Advantageously, the communication system 26 is also configured to receive the position of the dead reckoning navigation system 22 calculated by the dead reckoning navigation system 22 or by the relative positioning system 24 from the mobile terminal 29, the infrastructure 18 and / or the remote server.

[0087] Back to Figure 1 and Figure 2 As described above, the infrastructure 18 includes a plurality of beacons 54 disposed at access points 19 of the predefined space 14. These beacons 54 are typically grouped within a portal disposed at the access point 19, such as Figure 2 Each portal 60 includes at least three beacons 54 to allow pedestrians 12 passing through the access point 19 to be located by multilateral methods.

[0088] The infrastructure 18 is, for example, a permanent fixed infrastructure. This is particularly the case when the predefined space 14 is an industrial site and the positioning system 10 is intended to track the movements of technicians at said site. As a variant, the infrastructure 18 is a temporary fixed infrastructure. For example, when the predefined space 14 is an intervention site, in particular in the event of a fire: before the arrival of the firefighters, the infrastructure 18 is brought to an access point at the intervention site and installed. Again as a variant, the infrastructure 18 is a mobile infrastructure. It is usually loaded on a vehicle (not represented) for transporting pedestrians 12, whose movements towards the predefined space 14 are to be tracked. The vehicle itself is then equipped with a positioning and orientation system that makes it possible to calculate the position of the infrastructure 18 in a predetermined coordinate system. This is the case, for example, when the predefined space 14 is an intervention site, in particular in the event of a hostage situation.

[0089] Now refer to Figures 4 to 6 A method 100 implemented by the positioning system 10 , and more specifically by the processing units 32 , 52 , is described.

[0090] like Figure 4 As shown, the method 100 starts with a first step 102 of activating the dead reckoning positioning system 22. The first step 102 is usually performed when the pedestrian 12 wearing the box 16 is still outside the predefined space 14. Typically, the step 102 is triggered by the pedestrian 12 pressing a button (not shown) on the box 16. The relative positioning system 24 is usually started at the same time as the step 102.

[0091] After it is activated, the dead reckoning positioning system 22 provides the position of the pedestrian 12 in an arbitrary fixed coordinate system. The position v(t) is affected by random errors. For simplicity, the random errors are assimilated here to the central Gaussian variance variable For further simplification, it is subsequently assumed that there is no spatial correlation between the orientations of the planes, i.e., the term Covar(v 1 ,v 2 ) and Covar(v 2 ,v 1 ) is zero. For simplicity, let us consider the term Var(v 1 ) and Var(v 2 ) are equal, so the variance Γ v It can be written as where σ v (t) reflects the uncertainty of the position v(t), I 2 is an identity matrix of dimension 2. If this uncertainty σ v (t) is a time-varying a priori variable, the magnitude of this variation is usually low. Therefore, a constant uncertainty σ is considered here and below v .

[0092] The dead reckoning positioning system 22 also provides the heading ψ(t) of the pedestrian 12 in the arbitrary fixed coordinate system. This arbitrary fixed coordinate system is a two-dimensional horizontal coordinate system, which is generally composed of a rotation angle θ of a predetermined fixed coordinate system around a vertical axis and a translation of a horizontal vector Δ. Therefore, the arbitrary fixed coordinate system shares a common axis with the predetermined fixed coordinate system consisting of the vertical axis. It is arbitrarily selected by the dead reckoning positioning system 22 according to its orientation when it is started.

[0093] The relative positioning system 24 provides the horizontal position of the pedestrian 12 in a predetermined fixed coordinate system. The position u(t) is affected by random errors. For simplicity, this random error is assimilated here to a central Gaussian variance variable For further simplification, it is assumed in the following that there is no spatial correlation between the orientations of the planes, i.e., the term Covar(u 1 ,u 2 ) and Covar(u2 ,u 1 ) is zero. To simplify further, let us consider the term Var(u 1 ) and Var(u 2 ) are equal, so the variance Γ u It can be written as where σ u (t) reflects the uncertainty of the position u(t), I 2 is an identity matrix of dimension 2. If this uncertainty σ u (t) is a time-varying a priori variable, the magnitude of this variation is usually low. Therefore, a constant uncertainty σ is considered here and below u .

[0094] Step 102 is followed by step 104 for determining the likelihood of the relative positioning system 24 to estimate the position of the dead reckoning navigation system 22 in a predetermined fixed coordinate system. If the determination is positive, i.e. typically if the box 16 is within the range of the beacon 54 of the infrastructure 18, step 104 is followed by step 106 for resetting the position of the dead reckoning navigation system 22 and step 107 for resetting the heading of the dead reckoning navigation system 22. If the determination is negative, i.e. typically if the box 16 is out of the range of the beacon 54 of the infrastructure 18, step 104 is repeated after a time delay.

[0095] During the position resetting step 106, the dead reckoning navigation system 22 is at time t 0 Estimate the position v(t 0 ), and the relative positioning system 24 is at the time t 0 Estimate the position u(t 0 ), and the position of the evaluation u(t 0 ) is communicated to the dead reckoning navigation system 22. The dead reckoning navigation system 22 then determines a position reset parameter δ, which is determined by the estimated u(t 0 ) and the estimated v(t 0 ) The difference between the positions is: Δ = v(t 0 )-u(t 0 ). The dead reckoning navigation system 22 then adjusts the estimated position by applying the position reset parameter Δ:

[0096] refer to Figure 5 , when the box 16 is within range of the beacon 54 of the infrastructure 18 , a heading reset 107 begins with the pedestrian 12 following a movement trajectory 110 within a movement interval.

[0097] During the movement 110, for a plurality of evaluation moments i included in the movement interval k , the heading reset 107 includes at the evaluation time i k The relative positioning system 24 calculates the position u(i) of the pedestrian 12 in a predetermined fixed coordinate system. k ). After the evaluation 112, the dead reckoning navigation system 22 receives 114 the estimated position u(i k ).

[0098] At the same time, for the multiple estimated time points τ included in the moving interval k The heading reset 107 also includes the dead reckoning navigation system 22 in any fixed coordinate system at the estimated time τ k The position v(τ k ) is estimated at 116.

[0099] Following the receiving step 114 and the estimating step 116 is a step 118 in which the position u(i k ) and the estimated position v(τ k ) to match. In fact, the resetting step 107 is based on the use of an evaluation position that is synchronized with the estimated position. However, these two positions come from different sources, and the evaluation time i k Usually with the estimated time τ k Therefore, in evaluating u(i k ) and the estimated v(τ k ) positions need to be matched. This matching is intended to evaluate u(i k ) and the estimated v(τ k ) positions are associated and synchronized so that they can be determined at multiple times t k At each moment in the k The evaluated position u(t k ) and the estimated position v(t k ). Therefore, this matching includes the k The position u(i k ) of the set {u(i k )} k and from the estimated time τ k The estimated position v(τ k ) of the set {v(τ k )} k Inference:

[0100] - At a certain time t k The evaluated position u(t k ) of the set {u(t k )}k ,and

[0101] - At a certain time t k The estimated position v(t k ) of the set {v(t k )} k .

[0102] To this end, matching includes, for example, interpolating at least one set or subset among:

[0103] - At evaluation time i k The evaluation position u(i k ) of the set {u(i k )} k ,and

[0104] - at the estimated time τ k The estimated position v(τ k ) of the set {v(τ k )}k.

[0105] This interpolation preferably uses a temporal spline. Preferably, the spline used for the interpolation is at least twice differentiable. The spline representation can in particular enforce the continuity of the trajectory. The spline representation can also optimize the various parameters involved, in particular the temporal synchronization parameters, using a method based on the gradient of a criterion to be minimized. As a variant, the interpolation uses another method, such as, for example, a discrete representation of the trajectory followed by the pedestrian 12.

[0106] Preferably, only the set of estimated positions {v(τ k )}k interpolation to determine the time t k is chosen to be equal to the evaluation time i k :t k =i k Therefore, when determining the time t k The evaluated position u(t k ) of the set {u(t k )} k and at evaluation time i k The position u(i k ) of the set {u(i k )} k Therefore, the dead reckoning navigation system 22 receives 114 the position at the evaluation time i k The position u(i k ) constitutes the position of the pedestrian 12 at the determined time t obtained by the dead reckoning navigation system 22 k The evaluation position u(t k ) steps.

[0107] As a variation, only the set of evaluation positions {u(t k)} k Interpolate and determine the time t k is chosen to be equal to the estimated time τ k :t k =τ k Therefore, the match 118 constitutes the information obtained by the dead reckoning navigation system 22 about the pedestrian 12 at the determined time t k The evaluation position u(t k ) steps.

[0108] It should be noted that at the evaluation time i k Corresponding to the evaluation time τ k In this case, the match 118 is limited to the position u(i k ) and the estimated position v(τ k ) with the same moment i k , τ k are associated, and these moments then become the determined moments t k .

[0109] After the match 118, the matching pair of positions u(t k )、v(t k ) are added 120, and then the buffer memory 49 is cleaned 122. The cleaning 122 includes deleting from the buffer memory 49 the determined time t that is no longer in the movement interval. k The associated position pair u(t k )、v(t k ), where the moving interval is understood as a sliding time window of predetermined duration, at the most recent evaluation time i k Or the estimated time τ k Finish.

[0110] These steps 120, 122 are followed by a step 124 of inferring a heading reset parameter θ. This heading reset parameter θ here consists of a parameter, in particular an angle, for modifying the orientation by rotating around an axis of a predetermined coordinate system (here the vertical axis). Optionally, in parallel with step 124, the heading reset 107 also comprises other steps (not shown) of inferring other heading reset parameters, such as parameters for correcting deviations of trajectory angle changes.

[0111] refer to Figure 6 The step 124 includes a first sub-step 130 for calculating a first candidate value θ of the heading reset parameter θ 1 , where the buffer memory 49 contains all points, i.e., all position pairs u(t k )、v(t k ). The first candidate value θ 1is calculated by minimizing a cost function which is related to the position pairs u(t k )、v(t k ) corresponds to N determined moments t k The position u(t k ) and the N determined times t corrected by means of the heading reset parameter θ and the position reset parameter Δ k The estimated position v(t k ) for comparison. This cost function specifically represents the cost of applying the geometric transformation to the evaluation position {u k} k Afterwards, evaluate the position {u k} k and the estimated position {v k} k The average geometric deviation between the two, the geometric transformation includes:

[0112] - a rotation around an axis of a predetermined coordinate system (here the vertical axis) by an angle equal to the heading reset parameter θ, and

[0113] - A vector equal to the position resetting parameter Δ is translated in a direction orthogonal to the axis.

[0114] Therefore, the cost function is, for example, equal to in:

[0115] -{u k} k is the set of positions evaluated at a certain moment;

[0116] -{v k} k is the set of positions estimated at a certain moment;

[0117] -{t k} k It is a collection of definite moments;

[0118] N is the position pair u(t k )、v(t k )

[0119] -θ is the heading reset parameter;

[0120] - R(θ) is the rotation matrix representing the modified orientation by applying the heading resetting parameters; and

[0121] -Δ is the position reset parameter.

[0122] Therefore, the first candidate value θ 1 equal

[0123] in:

[0124] -u 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k a first position coordinate of the evaluation position along a first axis of a predetermined fixed coordinate system;

[0125] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on ;

[0126] -u 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k a second position coordinate of the evaluation position along a second axis of the predetermined fixed coordinate system;

[0127] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the second position coordinates on ;

[0128] -v 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k A first position coordinate of the estimated position along a first axis of an arbitrary fixed coordinate system;

[0129] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on ;

[0130] -v 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k A second position coordinate of the estimated position along a second axis of an arbitrary fixed coordinate system;

[0131] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the second position coordinates on ; and

[0132] N is the position pair u(t k )、v(t k ) quantity.

[0133] As a variant (not shown), the cost function consists of summing The weighting of different terms of {v} may be based on, for example, the likelihood of the terms (generally terms associated with higher variance will be given lower weights) or the geometric distance between different terms (generally terms that are geometrically close to each other will be given lower weights). As a variant, a geometric transformation is applied to the estimated position {v k} k , instead of evaluating the position {u k} k : Then the term in the summation is written as ‖u k -R(γ)v k -Δ‖ 2 As a variation, the geometric transformation also includes:

[0134] - trajectory smoothing, which is parameterized by the parameter b of the trajectory angle variation deviation correction: The summation term is then written as ‖v k -R(γ+b×t k ) k -Δ‖ 2 ;

[0135] and / or

[0136] - homothety, which is parameterized by the scaling factor h: The terms in the summation are then written as

[0137] ‖v k -h×(R(θ)u k +Δ)‖ 2 .

[0138] This sub-step 130 is followed by evaluating the first candidate value θ 1 The associated first precision value σ θ1 Sub-step 132. For example, the first precision value σ θ1 is the estimated position {v k} k and evaluate the position {u k} k The uncertainty σ v , σ u It is usually composed of the first candidate value θ 1 The standard deviation of is given by:

[0139] in:

[0140] -{u k} k is the set of positions evaluated at a certain time,

[0141] -{v k} k is the set of positions estimated at a certain time,

[0142] -{t k} k is a set of fixed moments,

[0143] - is the variance of the random error affecting each coordinate of the position estimated by the relative positioning system 24,

[0144] - is the covariance of the random errors affecting each coordinate of the position estimated by the dead reckoning navigation system 22,

[0145] -u 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k a first position coordinate of the evaluation position along a first axis of a predetermined fixed coordinate system,

[0146] -u 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k a second position coordinate of the evaluation position along a second axis of the predetermined fixed coordinate system,

[0147] -v 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the first position coordinate along the first axis of an arbitrary fixed coordinate system,

[0148] -v 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the second position coordinate along the second axis of an arbitrary fixed coordinate system,

[0149] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on

[0150] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the second position coordinates on

[0151] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on

[0152] - is the estimated position of the dead reckoning navigation system 22 at all determined times the average value of the second position coordinates on , and

[0153] N is the position pair u(t k )、v(t k ) quantity.

[0154] As a variant, the first precision value σ θ1 is the set of evaluation positions {u k} k and by means of having the first candidate value θ 1 The set of relevant estimated positions {v k} k is a function of the mean deviation observed between . For example, it is given by:

[0155] in

[0156] N is the position pair u(t k )、v(t k ),

[0157] -{u k} k is the set of positions evaluated at a certain moment;

[0158] -{v k} k is the set of positions estimated at a certain moment;

[0159] - is the set of dead reckoning navigation system 22 at a certain time The average of the evaluation positions on

[0160] - is the set of dead reckoning navigation system 22 at a certain time The average of the estimated positions on

[0161] -θ 1 is to assign the first candidate value θ 1 The heading reset parameters, and

[0162] -R(θ 1 ) is the rotation matrix representing the orientation modified by applying the heading reset parameters.

[0163] As a variant, the first precision value σ θ1 is the estimated position {v k} k and evaluate the position {u k} k The uncertainty σ v , σu function, and the set {u k} k With the help of having the first candidate value θ 1 The set of relevant estimated positions {v k} k A combination of functions of the mean deviations observed between them.

[0164] In parallel with sub-steps 130 and 132, the inference step 124 further includes a sub-step 134 of calculating a second candidate value θ of the heading reset parameter θ 2 , without the oldest point contained in the buffer memory 49, i.e., except for the oldest determined time t 1 The associated position pair u(t 1 )、v(t 1 ) has all the position pairs u(t k )、v(t k ). The second candidate value θ 2 is calculated by minimizing a cost function which takes into account the most recent position pair u(t k )、v(t k ) corresponds to the N-1 determined moments t k The position u(t k ) and the N-1 determined times t corrected by means of the heading reset parameter θ and the position reset parameter Δ k The estimated position v(t k ) for comparison. This cost function specifically represents the cost of applying the geometric transformation to the evaluation position {u k} k Afterwards, evaluate the position {u k} k and the estimated position {v k} k The average geometric deviation between the two, the geometric transformation includes:

[0165] - a rotation around an axis of a predetermined coordinate system (here the vertical axis) by an angle equal to the heading reset parameter θ, and

[0166] - A vector equal to the position resetting parameter Δ is translated in a direction orthogonal to the axis.

[0167] Therefore, the cost function is, for example, equal to in:

[0168] -{u k} k is the set of positions evaluated at a certain moment;

[0169] -{vk} k is the set of positions estimated at a certain moment;

[0170] -{t k} k It is a collection of definite moments;

[0171] N is the position pair u(t k )、v(t k )

[0172] -θ is the heading reset parameter;

[0173] - R(θ) is the rotation matrix representing the modified orientation by applying the heading resetting parameters; and

[0174] -Δ is the position reset parameter.

[0175] Therefore, the second candidate value θ 2 equal

[0176] in:

[0177] -u 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k a first position coordinate of the evaluation position along a first axis of a predetermined fixed coordinate system;

[0178] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time

[0179] The average value of the first position coordinates on ;

[0180] -u 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k a second position coordinate of the evaluation position along a second axis of the predetermined fixed coordinate system;

[0181] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time

[0182] The average value of the second position coordinates on ;

[0183] -v 1 (t k ) is the dead reckoning navigation system 22 at the determined time t kA first position coordinate of the estimated position along a first axis of an arbitrary fixed coordinate system;

[0184] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time

[0185] The average value of the first position coordinates on ;

[0186] -v 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k A second position coordinate of the estimated position along a second axis of an arbitrary fixed coordinate system;

[0187] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time

[0188] The average value of the second position coordinates on ; and

[0189] N is the position pair u(t k )、v(t k ) quantity.

[0190] As a variant (not shown), the cost function consists of summing The weighting of different terms of {v} may be based on, for example, the likelihood of the terms (generally terms associated with higher variance will be given lower weights) or the geometric distance between different terms (generally terms that are geometrically close to each other will be given lower weights). As a variant, a geometric transformation is applied to the estimated position {v k} k , instead of evaluating the position {u k} k : Then the term in the summation is written as ‖u k -R(θ)v k -Δ‖ 2 As a variation, the geometric transformation also includes:

[0191] - trajectory smoothing, which is parameterized by the parameter b of the trajectory angle variation deviation correction: The summation term is then written as ‖v k -R(θ+b×t k ) k -Δ‖ 2 ;

[0192] and / or

[0193] - isotope, which is parameterized by the scaling factor h: The terms in the sum are then written as ‖v k -h×(R(θ)uk +

[0194] Δ)‖ 2 .

[0195] This sub-step 134 is followed by evaluating the second candidate value θ 2 The associated second precision value σ θ2 Sub-step 136. For example, the second precision value σ θ2 is the estimated position {v k} k and evaluate the position {u k} k The uncertainty σ v , σ u It is usually composed of the second candidate value θ 2 The standard deviation of is given by:

[0196] in:

[0197] -{u k} k is the set of positions evaluated at a certain time,

[0198] -{v k} k is the set of positions estimated at a certain time,

[0199] -{t k} k is a set of fixed moments,

[0200] - is the covariance of the random errors affecting each coordinate of the position estimated by the relative positioning system 24,

[0201] - is the covariance of the random errors affecting each coordinate of the position estimated by the dead reckoning navigation system 22,

[0202] -u 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k a first position coordinate of the evaluation position along a first axis of a predetermined fixed coordinate system,

[0203] -u 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k a second position coordinate of the evaluation position along a second axis of the predetermined fixed coordinate system,

[0204] -v 1 (t k) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the first position coordinate along the first axis of an arbitrary fixed coordinate system,

[0205] -v 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the second position coordinate along the second axis of an arbitrary fixed coordinate system,

[0206] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time The average value of the first position coordinates on

[0207] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time The average value of the second position coordinates on

[0208] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time The average value of the first position coordinates on

[0209] - is the estimated position of the dead reckoning navigation system 22 at all determined times except the oldest time the average value of the second position coordinates on , and

[0210] N is the position pair u(t k )、v(t k ) quantity.

[0211] As a variant, the second precision value σ θ2 is the set of evaluation positions {u k} k and by means of having a second candidate value θ 2 The set of relevant estimated positions {v k} k is a function of the mean deviation observed between . For example, it is given by:

[0212] in:

[0213] N is the position pair u(t k )、v(t k ),

[0214] -{u k}k is the set of positions evaluated at a certain moment;

[0215] -{v k} k is the set of positions estimated at a certain moment;

[0216] - is the time determined by the dead reckoning navigation system 22 at all times except the oldest time The average of the evaluation positions on

[0217] - is the time determined by the dead reckoning navigation system 22 at all times except the oldest time The average of the estimated positions on

[0218] -θ 2 is to assign the second candidate value θ 2 The heading reset parameters, and

[0219] -R(θ 2 ) is the rotation matrix representing the orientation modified by applying the heading reset parameters.

[0220] As a variant, the first precision value σ θ2 is the estimated position {v k} k and evaluate the position {u k} k The uncertainty σ v , σ u function, and the set {u k} k With the help of having the first candidate value θ 2 The set of relevant estimated positions {v k} k A combination of functions of the mean deviations observed between them.

[0221] Sub-steps 132 and 136 are followed by the first precision value σ θ1 With the second precision value σ θ2 Sub-step 140 of performing a comparison.

[0222] If the first precision value σ θ1 The precision reflected is better than the second value σ θ2 Well, that is, here if σ θ1 <σ θ2 , then step 140 is followed by selecting the first candidate value θ 1 Step 142 as the selected candidate value.

[0223] On the other hand, if the accuracy value reflecting the best accuracy is given by the second accuracy value σθ2 If σ θ1 >σ θ2 , then step 140 is followed by step 144 of deleting the oldest point in the buffer memory 49, i.e., the oldest determined time t 1 The associated position pair u(t 1 )、v(t 1 ). Then the remaining determined moments Renumbered to And repeat steps 130 to 140.

[0224] This makes it possible to maximize the accuracy of resetting the parameters θ.

[0225] Sub-step 142 is optionally followed by a set of sub-steps 150 to 154 aimed at validating the new reset parameter θ in the residual calculation.

[0226] Sub-step 142 is followed by calculating the expected reset residual r exp Sub-step 150. The reset residual is intended to reflect the set of evaluation positions {u k} k and the set of related estimated positions {v k} k The expected mean deviation between . It is given by:

[0227]

[0228] in:

[0229] N is the position pair u(t k )、v(t k ),

[0230] - is the covariance of the random errors affecting each coordinate of the position estimated by the relative positioning system 24,

[0231] - is the covariance of the random errors affecting each coordinate of the position estimated by the dead reckoning navigation system 22,

[0232] - is the covariance of the reset parameter θ, equal to the selected candidate value θ 1 The standard deviation σ θ1 The square of, as mentioned above,

[0233] -u 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k a first position coordinate of the evaluation position along a first axis of a predetermined fixed coordinate system,

[0234] -u 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k a second position coordinate of the evaluation position along a second axis of the predetermined fixed coordinate system,

[0235] -v 1 (t k ) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the first position coordinate along the first axis of an arbitrary fixed coordinate system,

[0236] -v 2 (t k ) is the dead reckoning navigation system 22 at the determined time t k The estimated position of the second position coordinate along the second axis of an arbitrary fixed coordinate system,

[0237] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on

[0238] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the second position coordinates on

[0239] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the first position coordinates on , and

[0240] - is the estimated position of the dead reckoning navigation system 22 at all determined times The average value of the second position coordinates on .

[0241] Substep 150 is followed by calculating the observed reset residual r obs Sub-step 152. The reset residual is intended to reflect the set {u k} k and by assigning the selected candidate value θ 1 The set of relevant estimated positions {v k} k The average deviation observed between . It is given by:

[0242] in:

[0243] N is the position pair u(t k)、v(t k ),

[0244] -{u k} k is the set of positions evaluated at a certain moment;

[0245] -{v k} k is the set of positions estimated at a certain moment;

[0246] - is the set of dead reckoning navigation system 22 at a certain time The average of the evaluation positions on

[0247] - is the set of dead reckoning navigation system 22 at a certain time The average of the estimated positions on

[0248] -θ is the heading reset parameter, and

[0249] - R(θ) is the rotation matrix representing the modified orientation by applying the heading resetting parameters.

[0250] Sub-step 152 is followed by sub-step 154, where the observed reset residual r obs The expected reset residual r with a predetermined threshold δ added exp Make a comparison.

[0251] If the observed reset residual r obs Strictly greater than the expected reset residual r exp and the threshold δ, that is, if the inequality r is satisfied obs >r exp +δ, then sub-step 154 ​​is followed by sub-step 156 of rejecting the selected candidate value. The parameter θ is then reset to its previous value.

[0252] If the observed reset residual r obs Less than or equal to the expected reset residual r exp With threshold r exp The sum of, that is, if the inequality r is satisfied obs ≤r exp +δ, then sub-step 154 ​​is followed by sub-step 158, where the precision value σ associated with the selected candidate value is θn and the precision value σ associated with the current value of the reset parameter θ θo It should be noted that, considering the candidate value selection algorithm, the precision value σ associated with the selected candidate value θn Equal to the first precision value σ above θ1 .

[0253] More specifically, sub-step 158 includes associating the precision value σ with the selected candidate value θn and the precision value σ associated with the current value of the reset parameter θ θo By comparison, the reset parameter θ is increased by a time drift value proportional to the time elapsed between the determination of the selected candidate value and the determination of the current value of the heading reset parameter θ. The time drift value is generally equal to (t n -t o )×b, where:

[0254] -t n is the time instant {t k} k moment,

[0255] -t o is the determination time {t k} k moments, and

[0256] −b is a predefined constant representing a typical bias value of the gyroscope 40 .

[0257] At each moment t o ,t n For example, the determination time {t k} k As a variation, at each moment t o ,t n For example, the most recently determined time instant t considered for determining the selected candidate value and for determining the current value of the heading reset parameter θ, respectively k constitute.

[0258] In the case where sub-step 158 is first performed after the dead reckoning navigation system 22 is started, the accuracy value σ associated with the current value of the reset parameter θ θo Preferably, it is infinite.

[0259] Sub-step 158 is followed by sub-step 160 of determining a future value of the reset parameter θ based on the result of comparison 158. The future value is typically a function of the candidate value and the current value.

[0260] For example, if the precision value σ associated with the selected candidate θn Greater than or equal to the precision value σ associated with the current value θo , the current value increases the time drift value, that is, if the following inequality is satisfied: σ θn ≥σ θo +(t n -t o)×b, then the future value is equal to the current value, and on the other hand, if the precision value σ associated with the selected candidate value θn Strictly less than the precision value σ associated with the current value θo , the current value increases the time offset value, that is, if the following inequality is satisfied: σ θn <σ θo +(t n -t o )×b, then the future value is equal to the selected candidate value.

[0261] As a variant, according to the precision value σ associated with the selected candidate value θn and the precision value σ associated with the current value with the time drift value added θo The ratio between the future value and the current value is equal to the combination of the selected candidate value and the current value, usually by applying a Kalman filter or Bayesian fusion. For example, the future value is equal to in:

[0262] -θ n is the selected candidate value,

[0263] -σ θn is the precision value associated with the selected candidate value,

[0264] -θ o is the current value of the heading reset parameter θ, and

[0265] -σ θo is the precision value associated with the current value.

[0266] This sub-step 160 concludes the step 124 of inferring the heading reset parameter θ.

[0267] Return to Figure 5 , step 124 is followed by a step 126 of adjusting the estimated heading. During this step 126, the dead reckoning navigation system 22 adjusts the estimated heading by applying a heading reset parameter θ:

[0268] This step 126 ends the heading resetting step 107 .

[0269] Return to Figure 4 , the position resetting step 106 and the heading resetting step 107 are followed by a step 109 in which the dead reckoning navigation system 22 calculates the position of the pedestrian 12. During this step 109, the dead reckoning navigation system 22 applies the heading and position resetting parameters to determine the position of the pedestrian 12 in a predetermined coordinate system. The position is determined by the following formula: in:

[0270] - is the position of pedestrian 12 in the predetermined coordinate system calculated by dead reckoning navigation system 22,

[0271] - v(t) is the position of the pedestrian 12 in an arbitrary coordinate system estimated by the dead reckoning navigation system 22,

[0272] -θ is the heading reset parameter,

[0273] -R(θ) T is the rotation matrix that modifies the orientation by applying the inverse of the heading resetting parameter, and

[0274] -Δ is the position reset parameter.

[0275] The method 100 then loops back to step 109 to allow the calculated position to be Continuous updates.

[0276] At the same time, the method 100 returns to step 104 to allow for continuous updating of the heading reset parameter θ and the position reset parameter Δ each time the pedestrian 12 passes in the vicinity of the beacon 54 .

[0277] Thanks to the exemplary embodiments described above, the movement of people inside a building can be accurately tracked in a simple and economical manner. This goal is achieved thanks to the fact that only a particularly light infrastructure 18 needs to be deployed and only particularly simple positioning boxes 16 need to be equipped for the people being tracked. The motion sensors 30 equipped with these boxes 16 do not need to be very accurate, since the dead reckoning navigation system can be reset regularly each time a person passes near the infrastructure 18.

[0278] In particular, the use of UWB technology for the relative positioning system 24 is particularly advantageous because it allows a very accurate reset to be obtained in a small range of movement due to the low error level of UWB positioning. Therefore, it can significantly reduce infrastructure requirements. It is also completely transparent to the pedestrian 12, who does not have to perform any special tasks to ensure the reset of his positioning box 16.

Claims

1. A method (107) for resetting an attitude provided by a dead reckoning navigation system (22), the method (107) comprising the following steps: - estimating (116), by the dead reckoning navigation system (22), for each of a plurality of determined moments included in a movement interval in which the dead reckoning navigation system (22) moves along the movement trajectory, an estimated position of the dead reckoning navigation system (22) in an arbitrary fixed coordinate system at the determined moment, - obtaining (114) for each determination time instant an estimated position of the dead reckoning navigation system (22) in a predetermined fixed coordinate system at said determination time instant, said estimated position having been estimated by a relative positioning system (24), and - inferring (124) at least one pose reset parameter by minimizing a cost function comparing said evaluation position with said estimated position corrected by means of the or each reset parameter.

2. The posture resetting method (107) according to claim 1, wherein: The dead reckoning navigation system (22) is worn by a pedestrian (12), the dead reckoning navigation system (22) preferably being attached to a foot or ankle of the pedestrian (12).

3. The posture resetting method (107) according to claim 1 or 2, wherein: The dead reckoning navigation system (22) comprises a motion sensor (30) for measuring the motion of the dead reckoning navigation system (22) and a processing unit (32) for inferring the attitude and position of the dead reckoning navigation system (22) from the measured motion.

4. A posture resetting method (107) according to any one of the preceding claims, wherein: The relative positioning system (24) is selected from the group consisting of a multi-angle system, a multilateral system, a map matching system and a visual positioning system, and the relative positioning system (24) preferably includes an ultra-wideband telemetry device.

5. The posture resetting method (107) according to any one of the preceding claims, wherein: The arbitrary fixed coordinate system and the predetermined fixed coordinate system share a common axis, and the posture resetting parameter is composed of a parameter that modifies the direction by rotating around the common axis, and the parameter is preferably an angle.

6. A posture resetting method (107) according to any one of the preceding claims, wherein: The cost function represents the average geometric deviation between the evaluation position and the estimated position after applying a geometric transformation comprising a rotation and preferably a translation to the evaluation position or the estimated position.

7. A posture resetting method (107) according to any one of the preceding claims, wherein: The step of inferring the reset parameter comprises calculating (130) a candidate value for the reset parameter, evaluating (132) a precision value for the candidate value, comparing (158) the precision value with a previous precision value associated with a previous reset parameter, and determining (160) the reset parameter based on the result of the comparison (158), the reset parameter depending on the candidate value and the previous reset parameter.

8. The posture resetting method (107) according to claim 7, wherein: The accuracy value is a function of the uncertainty of the estimated position and the evaluated position and / or an average geometric deviation between the evaluated position and the estimated position after applying the candidate value of the reset parameter.

9. The posture resetting method (107) according to any one of the preceding claims, wherein: Inferring (124) the reset parameters comprises the following sub-steps: a) calculating (130) a first candidate value of the reset parameter by minimizing a cost function comparing the evaluated position at N determined moments with the estimated position at the N determined moments corrected by means of the reset parameter, -b) calculating (132) a first precision value associated with said first candidate value, -c) calculating (134) a second candidate value of the reset parameter by minimizing a cost function which compares the evaluation positions at the N-1 determination moments corresponding to the N determination moments minus the oldest determination moment with the position of the reset parameter corrected by means of the reset parameter or each reset parameter; The estimated positions at N-1 determined moments are compared. -d) calculating (136) a second precision value associated with said second candidate value, -e) comparing the first precision value and the second precision value (140), and -f) when the first precision value reflects the best precision, selecting (142) the first candidate value.

10. The posture resetting method (107) according to claim 9, wherein: When the precision value reflecting the best precision consists of the second precision value, deducing (124) the reset parameter comprises the following additional sub-steps: - deleting (144) the evaluated position and the estimated position at the oldest determined time instant, and - Repeating sub-steps a) to e), wherein the number N is reduced by 1.

11. A method (100) for locating an object (12) in a predefined space (14), the object (12) carrying a dead reckoning navigation system (22), the method (100) comprising the following steps: - starting (102) the dead reckoning navigation system (22), - resetting the attitude (107) of the dead reckoning navigation system (22) using a relative positioning system (24), The relative positioning system comprises an infrastructure (18) installed at an access point (19) of the predefined space (14) to obtain attitude reset parameters, the attitude reset (107) implementing the attitude reset method according to any of the preceding claims, - resetting the position (106) of the dead reckoning navigation system (22) using a relative positioning system (24), The relative positioning system comprises an infrastructure (18) installed at the access point (19) of the predefined space (14) to obtain position reset parameters, and - calculating (109) by the dead reckoning navigation system (22) a calculated position of the dead reckoning navigation system (22) in a predetermined coordinate system with the aid of the attitude reset parameters and the position reset parameters.

12. The positioning method (100) according to claim 11, wherein: The infrastructure (18) is carried on a vehicle which is itself equipped with a positioning and orientation system making it possible to calculate the position of the infrastructure (18) in the predetermined coordinate system.

13. A dead reckoning navigation system (22), comprising a motion sensor (30) for measuring the motion of the dead reckoning navigation system (22) and a processing unit (32) for inferring the attitude and position of the dead reckoning navigation system (22) in a fixed coordinate system from the measured motion, the processing unit being configured to implement the attitude resetting method (107) according to any one of claims 1 to 10 to reset the attitude.

14. A computer program product comprising code instructions for executing the posture resetting method (107) according to any one of claims 1 to 10 when the program is executed by a processor.

15. A storage device readable by a computer device, on which a computer program product is recorded, the computer program product comprising code instructions for executing the posture resetting method (107) according to any one of claims 1 to 10.

Images