FR3041769A1 - GEOLOCATION PROCESS - Google Patents

Abstract

The present invention relates to a method of geolocation of a mobile carrier (P), comprising steps of selecting (104) a bitter (A1) for the mobile carrier (P), measuring (106) a first distance (D1) between the bitter (A1) and the carrier (P), while the carrier (P) occupies a first position, calculating (110) a first estimate of the position of the bitter (A1) from of the first measured distance and an estimate of the first position of the carrier (P) provided by an inertial unit (INS), measuring (106) a second distance (D2) between the bitter (A1) and the carrier (P), while the carrier (P) occupies a second position different from the first position, calculating (110) a second estimate of the position of the bitter (A1) from the second distance (D2) and an estimate of the second position of the carrier (P) provided by the inertial unit (INS), correction (114) of a state of navigation of the carrier estimated by the central inertia lle (INS) from a difference between the first and second position estimates of the bitter (A1), so as to produce a corrected navigation state.

Description

FIELD OF THE INVENTION

The present invention relates to a method of geolocation of a mobile carrier.

STATE OF THE ART

To determine the absolute position of a mobile carrier, it is known to use an INS inertial unit coupled to a satellite navigation signal receiver (GNSS for example): this is called INS / GNSS hybridization.

However, in certain operational scenarios, the use of a constellation of satellites to ensure the accuracy of the inertial location is not possible (presence of jamming, luring, etc.). The accuracy of the inertial location can not be maintained and degrades.

Certain geolocation techniques operating in the absence of satellite positioning signals have thus been developed.

For example, methods known as Simultaneous Localization and Mapping (SLAM) are known, consisting in mapping the observed terrain and positioning the system in this reconstitution of the environment.

However, these methods require complex calculations in image processing. In addition, the SLAM method must be exploited continuously and requires the presence of many remarkable points in the observed images. Finally, this technique does not allow to have an absolute location of the system.

Processes are also known consisting in making positional readings of the wearer from the observation of a previously identified bitter.

A bitter is a fixed and unambiguously identifiable landmark used for navigation (a bell tower, a lighthouse, etc.) because it has certain structural features.

The predetermined absolute position of the bitter and the characteristic visual data of this bitter are stored in a database before the wearer's mission.

However, the database precisely locating the bitter that will then be exploited during the mission must be built before the beginning of the mission of the carrier: the operational mission is constrained by the passage in the vicinity of these predetermined bitter. Prior knowledge of the terrain is required.

SUMMARY OF THE INVENTION

An object of the invention is to geo-locate a mobile carrier in the absence of satellite positioning signals, without prior knowledge of the terrain.

It is therefore proposed, in a first aspect, a method of geolocation of a mobile carrier, comprising steps of: • selection of a bitter for the mobile carrier, • measurement of a first distance between the bitter and the carrier, while the carrier occupies a first position, • calculating a first estimation of the position of the bitter from the first measured distance and an estimate of the first position of the carrier provided by an inertial unit, Measuring a second distance between the bitter and the carrier, while the wearer occupies a second position different from the first position, calculating a second estimate of the position of the bitter from the second distance and of an estimation of the second position of the carrier provided by the inertial unit, • correction of a state of navigation of the carrier estimated by the inertial unit from a difference between the first and second position estimates of the bitter, so as to produce a corrected navigation state.

The method may be supplemented with the following optional features taken alone or in combination when technically possible.

The method may comprise: • a time stamp of the distance measurements between the bitter and the carrier, • the calculation of at least one speed error made by the inertial unit on the basis of the difference of position estimates of the bitter and timestamps.

The movable carrier can move from the first position to the second position along a bend path around the bitter.

The distances can be measured by a rangefinder whose line of sight is maintained towards the bitter during the movement of the carrier between the first position and the second position.

The aforementioned steps can be implemented for a first bitter and then repeated for a second bitter, and wherein the moving carrier moves in a zig-zag path between the two bitter. The correction step may be selectively implemented in response to the carrier detecting a satellite navigation signal loss.

The selection of the bitter may include the acquisition of an image by a camera of the wearer and identification of the bitter in the acquired image.

It is also proposed, in a second aspect, a geolocation system of a mobile carrier, comprising: • a device for selecting a bitter for the mobile carrier, • a rangefinder configured to measure: o a first distance between the l bitter and the wearer, while the wearer occupies a first position, o a second distance between the bitter and the wearer, while the wearer occupies at least a second position different from the first position, • a position estimator module bitter configured to calculate: o a first estimate of the position of the bitter from the first measured distance and an estimate of the first position of the carrier provided by a power plant, o a second estimate of the position of the bitter position; bitter from the second distance and an estimate of the second position of the carrier provided by the inertial unit, • a resetting module configured to correct a n state of navigation of the carrier estimated by the inertial unit from at least one difference between two estimates of position of the bitter.

This geolocation system can be included in an optronic viewfinder.

DESCRIPTION OF THE FIGURES Other features, objects and advantages of the invention will emerge from the description which follows, which is purely illustrative and nonlimiting, and which should be read with reference to the appended drawings, in which: FIG. schematic of a geolocation system of a carrier according to one embodiment of the invention. FIG. 2 is a flowchart of steps of a geolocation method according to one embodiment of the invention. • Figure 3 shows schematically the orientation of a carrier relative to the North. FIG. 4 shows an example of trajectory pursued by a wearer during the implementation of the method of FIG. 2.

In all the figures, similar elements bear identical references.

DETAILED DESCRIPTION OF THE INVENTION

Geolocation system

With reference to FIG. 1, a mobile carrier P, for example an aircraft, a ship or a land vehicle, comprises a geolocation system 1.

The geolocation system 1 can be part of an optronic viewfinder of the carrier P, this viewfinder having the main function of detecting and tracking targets.

The geolocation system 1 comprises an INS inertial unit.

The INS inertial unit includes inertial sensors such as accelerometers and gyrometers.

The INS inertial unit is configured to calculate an estimate of a navigation state of the mobile carrier P. The navigation state comprises one or more variables characterizing the movement of the carrier P (position, speed, attitudes, and corresponding errors).

The INS inertial unit typically comprises an extended Kalman filter EKF, that is to say that the inertial unit is configured to implement a Kalman filter type algorithm to calculate an estimate of the navigation state.

The geolocation system 1 furthermore comprises a satellite-based positioning signal receiver by a satellite constellation S, for example of the GPS type. This receiver is qualified in the GNSS receiver suite. In a particular embodiment, the GNSS receiver may be included in the inertial unit.

The GNSS receiver and the INS inertial unit are configured to form together an INS / GNSS hybrid power plant.

In such a hybrid power plant, the EKF extended Kalman filter of the INS inertial unit is configured to correct errors in a navigation state produced by the INS inertial unit, using measurements of the GNSS receiver.

The GNSS receiver is configured to detect a GNSS signal failure.

The Kalman EKF filter is also configured to correct a navigation state produced by the INS inertial unit using measurements provided by the rangefinder 8.

The geolocation system 1 also comprises a device 2 for detecting a bitter lying in view of the carrier P.

In the context of the present invention, a bitter is defined as a landmark fixed point in a geographical reference having singularities that can be detected visually or by other means. However, it is emphasized that the position of such a bitter is not a priori known to the wearer P.

The detection device 2 comprises, for example, a camera 4 and an image processing module 6 acquired by the camera 4. Alternatively or in a complementary manner, the detection device 2 comprises a difference-meter.

The geolocation system 1 also comprises a rangefinder 8.

The rangefinder 8 has a line of sight Y. The rangefinder 8 is, in a manner known per se, configured to measure along its line of sight Y a distance separating the rangefinder 8 (and therefore the carrier P which embeds it) and an object in view.

The geolocation system 1 further comprises orientation means 10 in the space of the line of sight Y of the geolocation system (central INS and rangefinder).

The geolocation system 1 is for example movably mounted relative to a support intended to be fixed to the body of the wearer P.

In a particular embodiment in which the carrier P is an aircraft, the support is fixed below the aircraft and the geolocation system 1 is mounted to move relative to the support by means of a hinge-type ball joint. In this way, when the aircraft is in flight, the geolocation system 1 can be oriented to any point on the ground.

In the embodiment illustrated in FIG. 1, the complete geolocation system 1 is mobile relative to the support fixed to the carrier P. It is however conceivable for the various components of the system 1 to be divided into two parts connected by the orientation means. 10. For example, it can be provided that the movable portion relative to the body of the carrier P includes only the camera 4 and the rangefinder 8. In this case, the orientation means 10 must deliver the relative positioning angles of the INS with respect to the line of sight of the rangefinder 8, with good accuracy.

The camera 4 is for example integral with the rangefinder 8 and positioned so that its optical axis X is parallel to the line of sight Y of the rangefinder 8.

The processing module 6 is configured to implement a target recognition in an image acquired by the camera 4 and a target tracking in a succession of images acquired by the camera using algorithms known from the state of the art. .

The processing module 6 is also configured to control the orientation means 10 of the rangefinder 8 and / or the camera 4.

The geolocation system 1 further comprises a bitter position estimator module 12, and a resetting module 14.

The estimator module 12 is configured to receive measurements acquired by the range finder 8 and to access an estimation of the navigation status provided by the inertial unit, and in particular a position estimation of the carrier P contained in this state.

The resetting module 14 communicates with the estimator module 12 and with the INS inertial unit.

The resetting module 14 is configured to correct a navigation state of the carrier estimated by the inertial unit (INS) from data provided by the bitter position estimator module.

The estimator module 12 and the resetting module 14 may be separate physical modules each comprising at least one processor for implementing calculations on the basis of data they receive.

In variants, these modules 12, 14 are virtual modules of a computer program capable of being executed by at least one processor.

The resetting module 12 and / or the estimator module 14 may, for example, be directly executed by at least one processor of the INS inertial unit or may form independent physical modules of the INS inertial unit.

Geolocation process

With reference to FIG. 2, a method of geolocation of the carrier P by the geolocation system 1 during the movement of the carrier P comprises the following steps.

The INS inertial unit produces 100 a state of navigation of the carrier P based on inertial measurements.

In the following, we consider a navigation state that includes at least one position information of the carrier, and his heading. By convention, the term "inertial position" is hereinafter called a position of the carrier P estimated by the inertial unit INS by means of its inertial sensors.

When at least one positioning satellite S is in view of the carrier P, preferably four satellites, the GNSS receiver receives positioning signals emitted by the satellite S, by which it develops and provides an estimate of its position at the Kalman filter. extended EKF. The navigation state produced by the inertial unit can then be corrected according to the positioning signals received using hybridization processing known per se.

The geolocation system 1 detects 102 at a given instant a loss of satellite positioning signal in the GNSS receiver. From this moment, the accuracy of the navigation status produced by the INS can not be corrected using the GNSS receiver.

This navigation state is then corrected in another way by implementing the steps below.

In a step 104, the detection device 2 detects a singular object A1 such as a bell tower. At this stage, the coordinates of this singular object A1 are not known from the geolocation system 1. The detection step 104 is typically implemented by means of the camera 4 and the processing module 6: the camera 4 acquires a image and communicates it to module 6; the latter analyzes the content of the image and detects the object A1 which is remarkable for certain visual characteristics (shape, contrast with the rest of the image, color, etc.).

The singular point A1 is used in the rest of the process as a bitter; we will designate it as such.

The processing module 6 generates and stores data representative of this bitter A1. The bitter A1 is detected on the fly during the movement of the carrier P, without the need for its coordinates to be determined before the start of the mission by the carrier P. In other words, no knowledge A priori mission field is not required for geolocation of the P-carrier.

In a step 106, the rangefinder 8 measures a distance D1 between the detected bitter A1 and the carrier P while the carrier P occupies a first position.

In a step 108 of time stamping, the measuring instant T1 of the measured distance D1 is moreover determined and stored by the geolocation system 1.

In a step 110, an estimate of the position of the bitter A1 is calculated by the estimator module 12 from the inertial position of the carrier P (supplied by the inertial unit INS by means of its inertial sensors) and the first distance. D1 measured by the rangefinder 8 between the carrier P and bitter A1.

With reference to FIG. 3, a first estimate of the position of the singular point A1 can for example be calculated as follows by the estimator module 12:

Pae (l) = Pie (l) + Tg / m (l) x DI where: • Pae (l) is the first positional estimate of bitter A1 at time T1 (for example a pair of coordinates (Xamer , Yamer), • Pie (l) is the inertial position of the carrier estimated by the INS inertial unit at time T1 (for example a pair of coordinates (Xins, Yins), • Tg / m (1) is a matrix of rotation image of the carrier's heading at time T1.

The carrier P moves 112 from the first position to a second position different from the first position.

Steps 106, 108, 110 are again implemented for the same bitter A1.

A new position estimate of bitter A1 is thus calculated as follows:

Pae (2) = Pie (2) + Tg / m {2) x D2 where: • D2 is the distance measured by the rangefinder in the second position, • T2 is the moment of measurement of the distance D2, • Pie ( 2) is the inertial position of the carrier estimated by the INS inertial unit at time T2, • Pae (2) is the estimated position of the bitter at time T2, • Tg / m (2) is a matrix of rotation image of the carrier's heading at time T2.

At this stage, therefore, there are two estimates Pae (1) and Pae (2) of the position of the same bitter Al.

During the movement of the carrier P between the first position and the second position, the camera 4 acquires a plurality of successive images of the bitter A1.

The analysis module 6 controls the orientation means 10 of the rangefinder 8 so as to maintain the line of sight Y of the rangefinder 8 towards the bitter A1, between the first position and the second position. This optional operation facilitates the implementation of the method because ensures that the object remains in view of the camera regardless of the path followed by the carrier P between the first position and the second position. The difference DP between the first and second estimates of position of the bitter (A1) produced by the estimator module 12 is written: DP = Pae (2) - Pae (l)

In a step 114, the resetting module 14 corrects a carrier navigation state estimated by the INS inertial unit from the DP difference between the first and second position estimates of the bitter A1, so as to produce a state corrected navigation.

In one embodiment, the registration 114 comprises the calculation of two different estimates of the distance traveled by the carrier P between the first position and the second position calculated by the inertial unit INS.

A first estimate of this distance traveled is calculated using the inertial positions provided by the INS inertial unit.

A second estimate of this distance traveled is calculated using the DP gap. The navigation state can then be corrected based on this comparison.

The sequence of steps 106, 108, 110, 112 may be repeated several times for the same bitter A1, so as to produce a plurality of estimates of the position of bitter A1 different Pae (1), Pae (2), Pae (3) .... The correction step 114 is also repeated, based on the difference between the last two estimates of bitter positions that have been calculated 110.

Preferably, the carrier P moves a T-turn trajectory around the bitter A1 between two successive distance measurements 106. In other words, the target object A1 is inside the turn.

Such a turn T trajectory is likely to more effectively correct the estimation errors made by the INS inertial unit by means of the resetting module 14.

The turn can for example turn 180 degrees around bitter A1.

The previously described steps in relation to bitter A1 are repeated for a new bitter A2.

A change of bitter is advantageously effected when the distance between the wearer P and the bitter being sighted exceeds a predetermined threshold.

This makes it possible to avoid the degradation of the measurements operated by the rangefinder 8.

For a good operation of the invention, the accuracy of the estimated position of the bitter is preferably less than 1m (from one measurement to another).

In one embodiment, the wearer moves along a trajectory T in zig-zag between two successive bitters A1, A2, such as that represented in FIG. 4.

Such a trajectory T in zig-zag makes it possible to benefit from the improvements of correction (thanks to turns successively left and right) while disturbing only moderately an optimal trajectory (typically substantially rectilinear) that would have followed the carrier P if no problem the receipt of GNSS signals had been encountered. The carrier P remains indeed confined in a corridor whose width depends on the radius of curvature of the turns and their lengths.

The transition from one bitter to another can be extended by a portion of straight line more or less long to reduce detours trajectory.

Correction example

Hereinafter, a simplified example of correction 114 according to one embodiment is described.

Let us consider two points of measurement of the position of a fixed bitter corresponding to 2 instants and 2 positions of the inertial unit.

Are:

Pav (1) = true position of bitter at time T1 Pav (2) = true position of bitter at time T2

Pae (1), Pae (2) = estimated position of the bitter at time T1, respectively T2 Piv (1), Piv (2) = true position of Tins at time T1, respectively T2 Pie (1) , Pie (2) = estimated position of the ins at time T1, respectively T2 Viv (1), Viv (2) = true speed of ins at time T1, respectively T2 Life (1), Life (2) = estimated speed of ins at time T1, respectively T2

First of all, to simplify the proof, let us consider that the speed error EV committed by the inertial unit is constant as a function of time, that is to say that EV = Old) - Viv (1) = Life ( 2) - Viv (2)

It is initially considered that the speed of the carrier is zero, that is

Viv = 0.

We have :

Pav (1) = Pivi 1) + Tg / mi 1) x Distance (1)

Pav (2) = Pivi2) + Tg / m (2) x Distance (2)

Paei 1) = Pie (l) + T g / m (l) x Distance i 1)

Pae (2) = Pie (2) + T g / miT ^ x Distancei ^) As the bitter darter is fixed, we also have:

Pavi 1) = Pau (2)

Pae (l) = Pae (2)

Let DP be the difference observed between the two successive measurements of the position of the bitter. DP = Pae (2) - Pae (l)

Moreover, since the carrier P moves, we have:

In the same way we have:

Now, we have Life = Viv + EV, the speed error EV committed by the inertial unit is constant and Viv = 0, so Vie = EV = constant.

It follows that :

Pie (2) - Pie (l) = EV x (Γ2 - 71)

Moreover, since the system has not moved,

Tg / m (1) = Tg / m (2)

Distance ^ 1) = Distance ^ 2)

Therefore,

Tg / m (l) x Distance ^!) = Tg / m (2) x Distance (2)

The calculation of the DP gap can then be rewritten as follows: DP = Pae (2) - Pae (l) = {Pie (2) + Tg / m (2) x Distance (2)} - {Pte (l) + Tg / m (l) x Distance (1)} = {Pie (2) - Pte (2)} + {Tg / m (2) x DistanceÇL) - Tg / m (Y) x Distance (l)} = EVX (T2-Tl)

It is thus demonstrated that the measurement of the positional deviation of the bitter DP makes it possible to go back to the speed error EV of the inertial unit INS.

The above can be generalized to a moving INS center; in this case, the difference in position of the plant will be compensated for by the difference in heading and distance during the measurement.

In practice, the trajectory T followed by the carrier P will directly influence the temporal behavior of the speed error.

By operating hypothesis, the speed is calculated in the geographical reference. It is calculated by integrating the acceleration into the geographical reference. The acceleration in the geographical reference is itself deduced by projecting the measurements of the accelerometers according to the heading of the sensors.

For simplicity, we have: Where:

• [g] is a variation in geographical speed • Acceleration [m] is an acceleration measured by the accelerometers, in the measurement frame. • Tg / m is the transition matrix from [m] to [g], cape image

The temporal behavior of the velocity error is deduced directly as a function of the heading error ECap.

Therefore, the speed error varies when the course of navigation is wrong and the speed varies.

For an airborne system, the easiest way to have a non-zero acceleration is to make a turn (in this case, the velocity vector is of constant standard, but of variable direction). If you have a turn, the trajectory is curved.

Furthermore, to optimally exploit what happens during this trajectory, one must be able to observe the variation of the speed error.

We have seen previously that it takes two measurements of the position of the bitter to estimate the speed error. It therefore takes 3 or 4 measurements to be able to estimate the variation of the speed error and thus to go back to the heading error.

Consequently, a curve-shaped, curve-shaped trajectory is advantageous in the context of the present invention because it makes it possible to go further than a speed error estimation; and in particular to go back to a heading error.

Claims (9)

1. A method of geolocation of a mobile carrier (P), comprising the steps of: • selecting (104) a bitter (A1) in view of the moving carrier (P), • measuring (106) a first distance (D1) between the bitter (A1) and the carrier (P), while the carrier (P) occupies a first position, • calculation (110) of a first estimate of the position of the bitter (A1) to from the first measured distance and an estimate of the first position of the carrier (P) provided by an inertial unit (INS), • measurement (106) of a second distance (D2) between the bitter (A1) and the carrier (P), while the carrier (P) occupies a second position different from the first position, • calculating (110) a second estimate of the position of the bitter (A1) from the second distance ( D2) and an estimate of the second position of the carrier (P) provided by the inertial unit (INS), • correction (114) of a carrier's sailing state estimated by the an inertial center (INS) from a difference between the first and second position estimates of the bitter (A1), so as to produce a corrected navigation state. 2. Method according to the preceding claim, further comprising: • a time stamp (108) of distance measurements between the bitter (A1) and the carrier (P), • the calculation of at least one speed error committed by the inertial unit based on the difference in bitter position estimates (A1) and time stamps. 3. Method according to one of the preceding claims, wherein the movable carrier (P) moves from the first position to the second position along a path (T) around the bitter bend (A1). 4. Method according to one of the preceding claims, wherein the distances are measured by a rangefinder (8) whose line of sight (Y) is maintained towards the bitter (A1) during the movement of the carrier ( P) between the first position and the second position. 5. Method according to one of the preceding claims, whose steps are implemented for a first bitter and repeated for a second bitter, and wherein the movable carrier (P) moves in a zig-zag path between the two bitter (A1, A2). The method according to one of the preceding claims, wherein the correcting step (114) is selectively implemented in response to the carrier's detection (102) of a satellite navigation signal loss. 7. Method according to one of the preceding claims, wherein the selection (104) of the bitter (A1) comprises the acquisition of an image by a camera (4) of the carrier (P) and an identification of the bitter (A1) in the acquired image. 8. Geolocation system for a mobile carrier (P), comprising: • a device (104) for selecting a bitter (A1) in view of the mobile carrier (P), • a rangefinder configured to measure: o a first distance (D1) between the bitter (A1) and the carrier (P), while the carrier (P) occupies a first position, o a second distance (D2) between the bitter (A1) and the carrier (P) , while the carrier (P) occupies at least a second position different from the first position, • a bitter position estimator module configured to calculate: o a first estimate of the position of the bitter (A1) from the first distance measured and an estimate of the first position of the carrier provided by an inertial unit (INS), o a second estimate of the position of the bitter (A1) from the second distance (D2) and an estimation of the second position of the carrier (P) provided by the inertial unit (INS), • a recalibration module configured to correct (114) a carrier navigation state estimated by the inertial unit (INS) from at least one difference between two position estimates of the bitter (A1). An optronic viewfinder comprising a system according to claim 8.