Method for determining an optimal trajectory for an aircraft carrying a ground observation sensor and product of associated computer program

Abstract

Method for determining an optimal trajectory for an aircraft carrying a ground observation sensor and associated computer program. This method consists of covering the rectangle encompassing an area to be observed (E) with a plurality of rectangular bars (106k) stacked edge to edge and optimizing a trajectory passing through a section of a base edge of each bar by: defining, for each bar, left (Pgk) and right (Pdk) points as the projections of the area to be observed covered by the bar in question; defining, for each bar, an active portion of the trajectory allowing the left and right points to be joined, and a passive portion of the trajectory allowing the right or left points to be joined to the next bar; and optimizing the trajectory by minimizing a cost function over all active and passive portions. Figure for the abstract: Figure 3

Description

Title of the invention: Method for determining an optimal trajectory for an aircraft carrying a ground observation sensor and associated computer program

[0001] The present invention relates to the field of methods for determining the trajectory that a surveillance aircraft must follow in order to sweep the entirety of a geographical area of ​​interest, by means of a ground surface observation sensor that this aircraft carries.

[0002] The trajectory is determined, for example, when the aircraft, while in flight, receives a new mission to perform, this mission indicating the area of ​​interest to be observed. The aircraft is then located at an origin point O and heads towards the area of ​​interest with an approach speed v, a speed that can be considered constant in direction and magnitude.

[0003] As illustrated in [Fig.4], the prior art method of determining the trajectory begins by delimiting a rectangle 5 which covers the area of ​​interest 4.

[0004] This rectangle, called the "enclosing rectangle", is oriented according to the approach speed of the aircraft. More precisely, the base side of the enclosing rectangle 5 is orthogonal to the approach direction D, a direction which is defined as the line parallel to the approach speed v of the aircraft and passing through the origin point O, the current position of the aircraft.

[0005] Then, the encompassing rectangle 5 is fully imaged by making the aircraft follow a trajectory T "in ladder", also called boustrophedon trajectory.

[0006] More precisely, it is assumed, as a first approximation, that the aircraft's altitude is constant. During the aircraft's movement, the sensor can then scan a strip on the ground having a predefined and constant width Lc.

[0007] The bounding rectangle 5, of length L and height H, is then tiled by a plurality of K strips 6k (k integer between 1 and K) of width Lc and oriented parallel to the base side of the bounding rectangle 5, two successive strips being arranged edge to edge, without mutual overlap.

[0008] The boustrophedon trajectory T, which is followed by the aircraft between a starting point Pin and an ending point Pout, allows the different bands to be swept successively.

[0009] The acquisition of the band of rank k is carried out by a displacement 7k of the aircraft parallel to the base side of the rectangle encompassing 5, in a first direction, for example from left to right on the [Fig.4].

[0010] Then the aircraft makes a 90° turn and moves 8k over a distance equal to the width Lc along a lateral side, for example right, of the bounding rectangle 5, before making another 90° turn.

[0011] The acquisition of the next band, of rank k+1, is carried out by a displacement 7k+i of the aircraft parallel to the base side of the rectangle encompassing 5 in a second direction, opposite to the first direction, for example from right to left on the [Fig.4].

[0012] However, this method is not optimal in terms of the time required to observe the entire geographical area of ​​interest. Indeed, it leads to scanning areas of the bounding rectangle 5 that are not part of the area of ​​interest 4. This unnecessarily lengthens the aircraft's trajectory and consequently the time required to complete the mission, i.e., to observe the entirety of the area of ​​interest alone.

[0013] The invention therefore aims to provide a method for determining the trajectory that allows the entirety of a predetermined geographical area to be swept while optimizing the time required for this operation.

[0014] To this end, the invention relates to a method, implemented by computer, for determining a boustrophedon trajectory for an aircraft carrying a ground observation sensor, characterized in that the method, for determining the trajectory allowing observation of the entirety of a predetermined area on the ground in a minimum of time, comprises the steps of: determining a convex polygonal envelope covering the area; constructing a candidate bounding rectangle for each of the sides of the envelope, one side of the candidate rectangle being superimposed on the side considered of the envelope; selecting, from the plurality of candidate bounding rectangles, the candidate bounding rectangle whose area is the smallest, as the bounding rectangle;cover the encompassing rectangle by means of a plurality of rectangular bars stacked edge to edge, a first bar of the stack resting on a so-called base side of the encompassing rectangle, the bars having a length equal to the length of the base side and a constant characteristic width, which depends on an altitude of the aircraft and the characteristics of the sensor;optimize a trajectory passing successively through at least one section of a so-called base edge of each of the bars, by defining, for each bar, a left point and a right point as the endpoints of a projection, onto the base edge of the bar in question, of the portion of the envelope covered by the bar in question, by defining, for each bar, an active portion of the trajectory allowing to join the left and right points defined on the base edge of the bar in question, by defining, for each bar, a passive portion of the trajectory allowing to join the points, right or left, of the bar in question and the next bar, and; by optimizing the trajectory by minimizing a cost function over all active and passive portions.

[0015] Depending on the particular embodiment, the process comprises one or more of the following characteristics, taken individually or in all technically possible combinations: - the sensor is a camera or a ground observation radar. - the trajectory is optimized for a constant aircraft altitude. - The sensor is designed to observe a segment on the ground with each step sampling, the sensor being mounted movable relative to the aircraft in order to be able to observe either a ground segment located to the right of the aircraft, or a ground segment located to the left of the aircraft, the orientation of the sensor relative to the aircraft not being changed along an active portion of trajectory. - along each of the active trajectory portions, the aircraft adopts a constant speed, preferably equal to a maximum speed allowed for optimal sensor operation. - a portion of passive trajectory to join a bar and the next bar includes a straight segment along the base edge of the bar in question or the next bar and a straight segment oriented at 90° allowing passage from the base edge of the bar in question to the base edge of the next bar. - each portion of passive trajectory is traversed with a speed whose amplitude is equal to that of the speed traversing the portions of active trajectory. - an active portion of the trajectory is traversed with the sensor on and a passive portion of the trajectory is traversed with the sensor off. - the optimization step is iterated by selecting, at each iteration, a different side of the encompassing rectangle as the base side against which to stack the plurality of bars, and thus obtain as many potential trajectories, the potential trajectory minimizing a travel time being selected as the optimal trajectory of the aircraft. - the optimization step implements a dynamic programming algorithm allowing to simultaneously consider a plurality of conceivable operational situations to link the acquisition of the active portion of one bar and the active portion of the next bar.

[0016] The invention also relates to a computer program product enabling the implementation of the previous process.

[0017] The invention and its advantages will be better understood upon reading the following detailed description of a particular embodiment, given solely by way of illustration and not limitation, the description being made with reference to the accompanying drawings in which:

[0018] - [Fig. 1] Fig. 1 is a schematic representation of the operational situation relating to the observation, by means of a sensor on board an aircraft, of an area of ​​interest on the ground;

[0019] - [Fig.2] [Fig.2] is a block representation of a mode of preferred embodiment of the process according to the invention;

[0020] - [Fig.3] [Fig.3] presents various illustrations of the implementation principle of the the process of [Fig. 2]; and,

[0021] - [Fig.4] Fig.4 is an illustration of the principle implemented by a process according to the state of the art.

[0022] Fig. 1 illustrates the operational situation in which a surveillance aircraft 1, carrying a sensor 2, makes an observation of the ground surface 3.

[0023] Sensor 2 is for example an optical camera, a radar, or any other type of sensor suitable for carrying out ground observation by scanning the ground surface.

[0024] The ground surface can be a sea surface, a land surface, or a mixed (land-sea) surface. The observation by the sensor can consist of detecting an object on the ground surface, or measuring a characteristic of the ground, or more generally any acquisition of information of interest.

[0025] The sensor 2 preferably has two modes of operation: an active mode, in which it is operational (or "on") and acquires information; and a passive mode, in which it is non-operational (or "off"), in the most general sense where it does not acquire information.

[0026] Aircraft 1 flies at an altitude h vertically above a point P, located on the ground. The Z-axis passing through aircraft 1 and point P is therefore the vertical axis. The X-axis is a longitudinal axis, parallel to the velocity v of aircraft 1, and the Y-axis is a transverse axis, perpendicular to the velocity of aircraft 1.

[0027] In active mode, at any given moment, the sensor 2 observes the ground 3 with an opening angle 0, enabling the acquisition of information relating to a ground observation segment. This observation segment is, for example, oriented along the transverse axis Y, between inner endpoint A and outer endpoint B. The distance between points A and B is the observation width Lc. The width Lc depends on the altitude of the sensor carrier, as well as the intrinsic and extrinsic characteristics of the sensor. At constant altitude and with constant characteristics of sensor 2, the width Lc is constant.

[0028] During the movement of aircraft 1, the movement of the observation segment AB parallel to the longitudinal direction X allows the scanning of a strip on the ground.

[0029] The interior point A of the observation segment AB may be different from point P. However, in the remainder of this description, for the sake of simplification and clarity, points A and P are superimposed, assuming that sensor 2 is properly oriented for this purpose. The trajectory in the X4 plane actually followed by aircraft 1 can then be approximated by the trajectory described by the interior point A. In general, however, the interval between points A and P should be taken into account to obtain the trajectory of aircraft 1 from the trajectory of point A.

[0030] The sensor 2 is movable relative to the aircraft 1 so as to have two positions allowing observation of an observation segment located to the right of the aircraft 1 or, symmetrically, an observation segment located to the left of the aircraft 1. The sensor 2 is oriented appropriately, then switched to active mode. Its orientation relative to the aircraft does not change while it remains in active mode.

[0031] Aircraft 1 carries a computer system 10.

[0032] The computer system 10 is a computer comprising computing means, such as a processor 11, storage means, such as a memory 12, and an input / output interface 13.

[0033] The input / output interface 13 is connected to a human-machine interface 14 comprising, for example, a screen 15 and input means 16 enabling an operator to monitor the progress of the observation mission. The input / output interface 13 is also connected to the sensor 2 in order to control the operating mode, to control the orientation of the sensor relative to the carrier aircraft, and to acquire the information collected by the sensor.

[0034] Memory 12 stores information such as the characteristics 18 of sensor 2 and the mission 19 assigned to the surveillance aircraft.

[0035] Memory 12 stores various computer programs, whose instructions, when executed by the processor 11, enable the implementation of different processes or functionalities.

[0036] Memory 12 stores in particular the instructions of a program whose execution enables the implementation of the method of determining an optimal trajectory according to the invention.

[0037] This program is referenced by the number 20 on [Fig. 1].

[0038] Figure 2 represents a particular embodiment of the method according to the invention. It corresponds to the execution of program 20 by the computer system 10.

[0039] The method 100 makes it possible to determine the optimal boustrophedon trajectory T that the aircraft 1 must follow so that the sensor 2 on it can scan the entire of a geographical area of ​​interest 4, on the ground, in a minimum amount of time. The optimization therefore focuses on the total duration of the mission.

[0040] The process 100 takes into account the following two constraints:

[0041] According to the first constraint, when sensor 2 is operating (active mode), its orientation is fixed, and aircraft 1 must maintain a constant speed in both direction and magnitude. This speed advantageously corresponds to the maximum speed that the aircraft can achieve while still allowing sensor 2 to correctly acquire information relating to the observation segment. As a result of this first constraint, aircraft 1 follows a straight trajectory at a constant speed when the sensor is in active mode. This trajectory segment, referred to as the active segment in what follows, extends between an activation point of sensor 2 (the sensor is switched on and enters active mode) and a deactivation point of sensor 2 (the sensor is switched off and enters passive mode).

[0042] The second constraint concerns the portions of the trajectory when the sensor 2 is in passive mode, referred to as passive portions in what follows. A passive portion allows two successive active portions of the trajectory T to be connected, that is, it connects the deactivation point corresponding to the exit end of one active portion and the activation point corresponding to the entry end of the next active portion of the trajectory T. For this purpose, the same function for evaluating the time required to connect two successive active portions is used throughout the optimization of the trajectory T.

[0043] For example, the evaluation function retained in the remainder of this description, which is ultimately similar to what is done in the prior art ([Fig.4]), assumes that the aircraft moves at a speed having a constant amplitude which is equal to that adopted by the aircraft when the sensor is in active mode; and that the passive portion allowing to join two successive active portions is made up of straight segments, at right angles to each other.

[0044] In a step 110, while aircraft 1 is located at origin point O, it receives a mission M to observe an area of ​​interest 4 on the ground.

[0045] Mission M is saved in a file 19 in the memory of computer system 10.

[0046] Among other information, mission M includes a description of area 4.

[0047] For example, zone 4 is a polygon, which is described by the list of coordinates geographical coordinates of the vertices of this polygon.

[0048] Mission M also defines the altitude h at which aircraft 1 should preferably maintain itself in order to carry out the observation of the entire zone 4.

[0049] In step 120, the width Lc of the observation segment that the sensor 2 can image at each instant is calculated, from the altitude h and the known characteristics of the sensor 2, such as its aperture, orientation, etc., this information being stored in the memory of the computer system 10.

[0050] As previously stated, to simplify the present description, the observation segment is considered to extend from the vertical through aircraft 1, either to its right or to its left, with sensor 2 being properly oriented before switching into active mode and not changing orientation as long as it remains in active mode.

[0051] Thus, during an activation period of the sensor 2, it acquires a strip on the ground having a width Lc and a length equal to the product of the speed of the aircraft 1 by the duration of the activation period.

[0052] Step 130 consists of defining the convex polygonal envelope E of zone 4 from the description of zone 4.

[0053] If zone 4 is already convex, the envelope E corresponds to the perimeter of zone 4.

[0054] The method of determining such an envelope is known and is outside the scope of the present invention.

[0055] The convex polygonal envelope E has N vertices and N sides. The side of rank n connects the vertices of rank n and n+1 (n is an integer between 1 and N).

[0056] The next step, 140, consists of defining the rectangle encompassing the envelope E which has the smallest area among the set of possible encompassing rectangles.

[0057] To achieve this, we rely on a result from geometry according to which the area of ​​the bounding rectangle associated with the nth side of a polygon is minimal when this bounding rectangle rests on this nth side. In other words, if we define the orientation of the bounding rectangle associated with the nth side by the angle r between the base side of this bounding rectangle and the side of rank n of the envelope (as illustrated in [Fig. 4]), the area of ​​the bounding rectangle is minimal when the angle r is zero.

[0058] Step 140 thus begins with the construction (step 142) of a set of N candidate bounding rectangles.

[0059] Each candidate rectangle rests on a different side of the envelope E.

[0060] More precisely, the candidate bounding rectangle of rank n, Rn, is obtained by selecting the side of rank n of the envelope E.

[0061] A baseline is defined as the line passing through the side of rank n.

[0062] The next step is to erect heights, perpendicular to the baseline, and originating from the vertices of rank n and n+1 of the envelope E.

[0063] Then, these heights are moved along the baseline so that the surface delimited between these two heights covers the envelope E.

[0064] The minimum distance between these two heights allowing the envelope E to be covered gives the length of the base side of the candidate bounding rectangle Rn. This length is denoted Ln.

[0065] Then, the position of the side opposite the base side is determined. For this, the opposite side is moved parallel to the base side so that the surface delimited between the base side and the opposite side covers the entire envelope E.

[0066] The minimum distance between the base side and the opposite side allowing the envelope E to be covered gives the height of the candidate encompassing rectangle Rn. This height is denoted Hn.

[0067] Finally, the area Sn of the nth encompassing candidate rectangle Rn is determined by multiplying its length by its height: Sn = Ln x Hn.

[0068] Once the areas of the N candidate bounding rectangles have been determined, the candidate with the smallest area Sn is selected (step 144) as the bounding rectangle R for the continuation of process 100.

[0069] In step 150, a plurality of bars are defined so as to cover the rectangle encompassing R.

[0070] In step 152, the side of the encompassing rectangle R against which the bars are "stacked" is selected. For example, the base side of rectangle R is selected (defined as the side of rectangle R closest to point O).

[0071] Then, in a step 154, the way of covering the rectangle encompassing R based on the side selected in step 152 is determined.

[0072] The number of bars is equal to K. The rank of each bar is indexed by the integer k, between 1 and K. The bars are referenced 106k on the [Fig.3].

[0073] A bar has a rectangular shape. Each bar has a width equal to the width Lc determined in step 120. The length of each of the bars 106k is equal to the length of the side selected in step 152.

[0074] The 106k bars are arranged edge to edge, without overlap.

[0075] The first bar 106i rests, by its so-called base side, on the base side of the rectangle encompassing R. The second bar 1062 rests, by its so-called base side, on the opposite side of the first bar 106i, etc.

[0076] Since the bars 106k have a predefined width Lc, the stack of K bars can have a dimension slightly greater than the height H of the enclosing rectangle R. In other words, the opposite side of the enclosing rectangle lies inside or on the opposite side of the bar of rank K. K is in fact equal to the integer part of the ratio of the height H of the enclosing rectangle divided by the thickness Lc, plus one. For example, K is equal to 6 in [Fig. 3].

[0077] The output of step 150 is illustrated in diagram A of [Fig. 3]. In the example shown, zone 4 is convex. It corresponds exactly to its envelope E. The bounding rectangle R rests on one of the sides of the zone of interest 4. The bounding rectangle R is covered by six bars 6k of width Lc, “stacked” on the base side of the bounding rectangle.

[0078] The process 100 continues with step 160 of determining the optimal trajectory T. This is a boustrophedon trajectory which follows, at least in part, the base sides of the different bars 106k.

[0079] This step is based on the following observations. If aircraft 1 were to move along the entire base side of a bar, such as bar 106k in Figure 3A, its path could be broken down into several phases: - a first passive phase q>i, of movement of aircraft 1 while sensor 2 is off. This first phase extends from one end (for example right) of the base side of the bar, until reaching an activation point, Pdk, corresponding to the projection of the portion of the envelope E covered by the bar 106k onto the base side of the bar. - a second active phase q>2, of movement of aircraft 1 while sensor 2 is activated so as to image a part of the 106k bar corresponding to a band scanned by the sensor. This second phase extends between the activation point Pdk and a deactivation point Pgk, the latter corresponding to the projection, on the base side of the 106k bar, of the portion of the envelope E covered by the 106k bar. - a third passive phase, q>3, of displacement of aircraft 1, sensor 2 off, from the deactivation point Pgk to the other end (e.g. left) of the base side of the 106k bar. - a fourth passive phase, q>4, of displacement of aircraft 1 sensor 2 off, along a lateral edge (for example the left edge) of the bar 106k so as to join the base side of the next bar.

[0080] The second phase is optimal as such. Indeed, it depends only on the parameters of the sensor, in this case the parameter Lc and the optimal speed that aircraft 1 must adopt for the acquisition of information by the sensor.

[0081] The fourth phase is optimal as such. According to the second constraint indicated above, the speed of the aircraft is considered to be the optimal speed.

[0082] On the other hand, the first and third phases can be optimized jointly. Indeed, the third phase of the k-th bar is linked with the first phase of the k+1-th bar, and this with an opposite direction of traversal (trajectory (in boustrophedon). To shorten this sequence to the bare minimum, it is possible to remove either the third phase for traversing the k-rank bar or the first phase for traversing the k+1-rank bar.

[0083] Thus, for each bar 106k, the intersection of the envelope E with the bar 106k is projected onto the base edge of the bar 106k. The endpoints of this projection define a left point of rank k, Pgk, and a right point of rank k, Pdk.

[0084] Then, for each rung 106k, the sensor-activated path between the right and left points of rank k, Pgk and Pdk, defines the second phase for rung of rank k. The k-rank strip corresponds to the portion of rung of rank k between points Pgk and Pdk observed by sensor 2 while aircraft 1 travels along the base side segment of rung 6k between points Pgk and Pdk. Thus, the active portion of the trajectory connecting the left and right points is defined.

[0085] The output of this procedure is illustrated in diagram B of [Fig.3]. For each bar 106k, the left and right endpoints are defined which delimit each active portion of the trajectory T.

[0086] It is then necessary to determine how to link the sweeping of the different bands associated with each of the bars, in order to minimize overall the cumulative length of the first and third phases, that is to say to define for each bar 106k, the passive portion of trajectory allowing to join the points, right or left, of the bar in question and the next bar.

[0087] An optimization of the overall trajectory is achieved by minimizing a cost function over all active and passive portions

[0088] Advantageously, this optimization is implemented by means of a dynamic programming algorithm which allows for the simultaneous consideration, for each bar of rank k, of four possible operational situations: - first situation, traverse from left to right of the bar of rank k and join the bar of rank k+1 from the left; - second situation, traverse from left to right of the bar of rank k and join the bar of rank k+1 from the right; - third situation, traverse the bar of rank k from right to left and join the bar of rank k+1 from the left; and, - fourth situation, traverse the bar of rank k from right to left and join the next bar of rank k+1 from the right.

[0089] The dynamic programming algorithm is advantageously initialized with the length of the path required to join the origin point O with an entry point Pin which is chosen as being either the left point Pgi of the rank 1 band, or the right point Pdi of the rank 1 band.

[0090] It should be noted that at level k+1 of the calculation tree (corresponding to the consideration of the k+1 rung), the dynamic programming algorithm considers the pair of cases from the previous level k (considering the k rung), namely on the one hand the length of the path exiting from the left and on the other hand the length of the path exiting from the right, and produces, for each of these cases for level k, a new pair of cases allowing the optimal crossing of the k+1 rung. Once the entire tree has been calculated, the fastest trajectory, i.e., the one whose cumulative length of the first and third phases is the shortest, is finally selected.

[0091] The output of step 160 gives the optimal trajectory T allowing to minimize the time required to acquire the area of ​​interest 4 by the sensor 2.

[0092] The exit of step 160 is illustrated in diagram C of [Fig.3]. The trajectory T is minimal in that the passive portions of the trajectory (each composed of a straight segment parallel to the base side of the rectangle encompassing R and a straight segment oriented at 90° to the base side of the rectangle encompassing R and allowing the two base sides of successive bars to be connected) are reduced in length.

[0093] Advantageously, steps 150 and 160 are iterated by selecting another side of the bounding rectangle R as the base side against which to stack the bars to cover the bounding rectangle. Each iteration then leads to the calculation of a trajectory indexed by the integer j, 1).

[0094] In this case, in a step 170, the shortest trajectory Tj is selected as the trajectory T to be realized.

[0095] This trajectory is defined, for example, by: - the entry points Pin and exit points Pout of the trajectory T and therefore of the execution of the mission, as well as the way to traverse each portion of the bar; - a first partial length corresponding to the total length of the path of the sensor turned on; - a second partial length corresponding to the total length of the path of the sensor off.

[0096] The method just described provides the optimal trajectory. Moreover, in terms of computational load, the proposed algorithm is quite lightweight and can therefore be executed by a computer on board an aircraft, in particular a light aircraft such as a drone, and having limited computing power.

Claims

Demands

1. A computer-implemented method (100) for determining a boustrophedon trajectory (T) for an aircraft (1) carrying a ground observation sensor (2), characterized in that the method, for determining the trajectory allowing observation of the entirety of a predetermined area on the ground in a minimum time, comprises the steps of: - determine (130) a convex polygonal envelope (E) covering the area; - construct (142) a candidate bounding rectangle for each of the sides of the envelope, one side of the candidate rectangle being superimposed on the considered side of the envelope; - select (144), from the plurality of candidate bounding rectangles, the candidate bounding rectangle whose area is the smallest, as bounding rectangle (R); - cover (150) the encompassing rectangle by means of a plurality of rectangular bars (106k) stacked edge to edge, a first bar of the stack resting on a so-called base side of the encompassing rectangle, the bars having a length equal to the length of the base side and a constant characteristic width, which depends on an altitude of the aircraft and the characteristics of the sensor; - optimize (160) a trajectory passing successively through at least one section of a so-called base edge of each of the bars, by: • defining, for each bar (106k), a left point (Pgk) and a right point (Pdk) as the endpoints of a projection, onto the base edge of the bar considered, of the portion of the envelope covered by the bar considered; • defining, for each bar (106k), an active portion of trajectory allowing to join the left and right points defined on the base edge of the bar considered; • defining, for each bar (106k), a passive portion of the trajectory allowing it to reach the points, right or left, of the bar in question and the next bar; and, • optimizing the trajectory by minimizing a cost function over all active and passive portions.

2. A method according to claim 1, wherein the sensor (2) is a ground observation camera or radar.

3. A method according to claim 1 or claim 2, wherein the trajectory (T) is optimized for a constant altitude (h) of the aircraft (1).

4. A method according to any one of the preceding claims, wherein the sensor (2) is adapted to observe a ground segment at each sampling step, the sensor being mounted movable relative to the aircraft (1) so as to be able to observe either a ground segment located on the right of the aircraft, or a ground segment located on the left of the aircraft, the orientation of the sensor relative to the aircraft not being changed along an active trajectory portion.

5. A method according to any one of the preceding claims, wherein, along each of the active trajectory portions, the aircraft (1) adopts a constant speed, preferably equal to a maximum speed allowed for optimal operation of the sensor.

6. A method according to any one of the preceding claims, wherein a portion of a passive trajectory for joining a bar and the next bar comprises a straight segment along the base edge of the bar in question or the next bar and a straight segment oriented at 90° allowing passage from the base edge of the bar in question to the base edge of the next bar.

7. A method according to claim 5 or claim 6, wherein each portion of the passive trajectory is traversed with a speed whose amplitude is equal to that of the traversal speed of the active portions of the trajectory.

8. A method according to any one of the preceding claims, wherein an active portion of the trajectory is traversed with the sensor on and a passive portion of the trajectory is traversed with the sensor off.

9. A method according to any one of the preceding claims, wherein the optimization step is iterated by selecting, at each iteration, a different side of the encompassing rectangle (R) as the base side against which to stack the plurality of bars, and thus obtain as many potential trajectories as possible, the potential trajectory minimizing a travel time being selected (170) as the optimal trajectory of the aircraft.

10. A method according to any one of the preceding claims, wherein the optimization step implements a dynamic programming algorithm that allows for the simultaneous consideration of a plurality of possible operational situations for chaining the acquisition of the active portion of one bar and the active portion of the next bar.

11. Product computer program comprising software instructions which, when executed by a computer system (10), implement a method (100) according to any one of the preceding claims.

Images