SYSTEM FOR DETECTING THE PATH OF MOVING OBJECTS

Description

Scope of the invention

[0001] 1. The field of the invention relates to the detection of objects in space at low and high altitudes, using an array of telescopes. More particularly, the field of the invention relates to the detection of satellites and space debris, as well as aircraft or any moving objects illuminated by the sun and the observer at night, and to the calculation of their orbits and trajectories in order to prevent them from falling to Earth, colliding in space, or entering a danger zone.

[0002] 2. Detecting orbital debris is very problematic. Indeed, tracking an object whose size may be less than ten centimeters, at a distance of several thousand kilometers, is an arduous task.

[0003] 3. On the other hand, since objects have very different trajectories from each other, it is very difficult for economic reasons to cover the entire celestial vault with adequate detection capabilities.

[0004] 4. The ESA (European Space Agency) estimated the number of active and inactive satellite space objects, bodies abandoned during satellite launches, rocket fairings, rocket bodies, and small fragments at: 5,400 pieces of space debris larger than one meter; 34,000 pieces of space debris larger than 10 centimeters; 900,000 pieces of space debris larger than one centimeter; 130,000,000 pieces of space debris larger than one millimeter.

[0005] 5. Currently, monitoring near-Earth space and in particular objects presenting a potential danger, whether natural or artificial, has become a major problem to ensure the safety and integrity of any target that may be involved in a collision.

[0006] 6. This monitoring concerns both active, end-of-life or uncontrolled satellites and debris from previous collisions, wear and tear on objects placed in orbit or asteroids or comets potentially dangerous to Earth.

[0007] 7. Space agencies and private operators have initiated various monitoring programs grouped under the name "Space Situational Awareness".

[0008] 8. Even when a meteorite is relatively small, the risks posed by a fall remain significant. A danger persists for populations and infrastructure on Earth's surface from any falling object, as its fragmentation in the atmosphere can cause substantial damage. One of the challenges is cataloging small bodies in the solar system whose orbits intersect Earth's and tracking them to assess their potential hazard.

[0009] 9. There is a growing number of artificial objects in low and high Earth orbits. Space congestion appears to be increasing every year due to the growing number of satellites and the increasing frequency of catastrophic collisions, resulting in a greater amount of debris. Poor or nonexistent tracking of their trajectories remains a growing problem.

[0010] 10. Projections suggest a situation in which the increasing population of debris larger than 1 cm will lead to increasingly difficult control and monitoring. The 1 cm threshold corresponds to the size of an object that could potentially render a satellite partially or completely inoperable, given the speeds involved: 3 km / s in geostationary orbit up to 8 km / s in low Earth orbit.

[0011] 11. We will call these objects "space objects" in the rest of the text knowing that this notion includes debris proper, operational or non-operational satellites, or even meteorites.

[0012] 12. The first problem concerns debris falling to Earth's surface, and the second concerns collisions between debris and active satellites. Monitoring debris in space, particularly in low Earth orbit, helps prevent both of these consequences.

[0013] 13. Finally, the issue of monitoring various space-based objects also extends to discreet objects moving at very low altitudes, such as aircraft (for example, ultralight aircraft or drones), which can pose a danger, for instance, when they fly near a sensitive site. One challenge is finding a wide-field optical system capable of covering a significant portion of the sky with sufficient resolution to detect objects at different altitudes, both near and far, and to track low-altitude objects with high speeds that are therefore difficult to detect.

[0014] 14. Indeed, one problem with the detection and monitoring of space debris, whose orbit and / or trajectory is not known a priori, is taking into account the light intensity of third-party sources which disrupts detections.

[0015] 15. These sources can originate from the celestial vault, the sun, the moon, and local weather conditions that alter the stability of the image capture conditions. The surveillance system must be able to take into account a multitude of lighting conditions to maximize detection under all circumstances. Since detection is performed by considering a point or area on the Earth's surface, the observer's field of view is an extremely important factor in calculating the probability of detecting a moving object and determining its trajectory.

[0016] 16. The issue of debris monitoring involves considering different orbits in space object detection methods. For natural objects such as meteorites, their orbit is generally heliocentric, meaning they can potentially approach Earth at any altitude and from any direction. For artificial objects, their Earth orbit can fall into different orbital categories.

[0017] 17. The first family of orbits is known by the acronym LEO, which stands for "Low Earth Orbit" in English. It is a family of low Earth orbits extending up to 2000 km. This family of orbits is heavily used by communications, military, detection, meteorological, and other satellites.

[0018] 18. A second family of orbits is known by the acronym GEO, standing for "Geostationary Orbit," which includes a geostationary orbit defined at 35,784 km above the equator. A revolution of a moving object at this altitude takes 24 hours. An object in a geostationary orbit is fixed relative to a position on Earth. However, debris can leave its orbit and follow orbits that are not directly above the Earth. This orbit is heavily used by communication satellites (military or civilian), remote sensing satellites, meteorological satellites, etc.

[0019] 19. A third family of orbits is better known by the acronym MEO, which stands for "Medium Earth Orbit" and is a family of medium orbits, generally elliptical. This is the case for GNSS satellites.

[0020] 20. A fourth family of orbits is designated by the acronym HEO, highly elliptical orbits, including very elliptical orbits such as Molniya or Tundra orbits, which allow communication or monitoring of high latitude regions.

[0021] 21. A fifth family of orbits is designated by the acronym GTO, meaning "Geostationary Transfer Orbit." This family comprises elliptical orbits. Their apogee is approximately 42,000 km and their perigee is approximately 650 km. This family of orbits is very practical for injecting satellites into a geostationary orbit; it is therefore used during satellite launches as a transition orbit for geostationary placement.

[0022] 22. Today, various methods exist for detecting space debris and its trajectory.

[0023] 23. In particular, there is a family of methods known as "active methods," notably for detecting debris in LEO (low Earth orbit). Active methods rely on a radar-like operation in which a moving object is illuminated by a signal-emitting source. The signal is then reflected, and it is this reflection that informs a receiver of the object's position data.

[0024] 24. A primary drawback of this method is that the received power varies as 1 / d⁴, where "d" is the distance from the mobile device to the transmitter / receiver. Consequently, the received flux will remain low during detection, even if a high transmission power is considered.

[0025] 25. A second drawback is the relatively large installation of radar-type systems that this method requires. These installations are expensive, require significant maintenance, and are easily detectable. Furthermore, these systems consume a great deal of energy and therefore must be installed near a power grid.

[0026] 26. Among active methods, there are also LiDAR systems, which rely on illuminating a moving object with a laser. This method achieves better results than radar in terms of detected power because the laser light is more focused. However, the detection cones are much weaker and are poorly suited to "blind" detection of objects in low and elliptical orbits. Another family of methods exists: passive methods. When applied to radar, these methods involve only receiving equipment. They must therefore be located near a powerful radar emission source. Detection technologies in which targets are not illuminated by a terrestrial source are also considered passive.Regarding passive methods, the luminous flux captured by a detector varies with the distance "d" from the moving object as a function of 1 / d², which offers better results than active methods for the luminous flux captured from the moving object. However, the major drawback is the strong dependence on illumination from external sources such as the sun, stars, or the moon. The advantage of these solutions lies in their low cost and the relative simplicity of their implementation using detectors based on optical instruments capable of visualizing small objects at all altitudes.

[0027] 27. With regard to objects in geostationary orbit, a telescope, radar, or any other electromagnetic instrument can detect a stationary point against a background of moving stars during the exposure time. With a wide-field telescope, it is then possible to detect space objects in a geostationary orbit and their trajectories.

[0028] 28. In the case of other orbits, called "sliding" and therefore non-geostationary, that is to say not fixed with respect to an observation point on Earth, it is difficult to detect space mobiles, their height, their inclination, the different nodes of their orbit, including the ascending and descending nodes.

[0029] 29. One difficulty arises from the speed at which spacecraft move, which can exceed 1° / s at the zenith for a low Earth orbit. Detection is achieved by capturing a trace (in a sequence of images) relative to point traces or trails, depending on the sidereal motion and therefore the observation window in the sky.

[0030] 30. The method then consists of discriminating between tracks to detect the presence of space debris. While the inclination can potentially be detected by analyzing the track left by the object, obtaining the object's actual altitude remains very difficult due to the unknown distance. However, it can be estimated from the object's velocity. Consequently, it is difficult to deduce elements of its trajectory by extrapolating from track analyses. In general, three angular position measurements of the object are necessary to derive its orbit. Two measurements are sufficient if the object is in a circular orbit.

[0031] 31. The problem can be solved by increasing the field of view of a telescope to increase the traces and their number, but the images detected, as explained previously, can become difficult to analyze due to the number and complexity of the telescopes to be implemented, the surrounding light pollution, the strong confusion caused by all the objects in the field, and the very large size of the sensors required.

[0032] 32. Moreover, the construction of a wide-field telescope is hardly conceivable beyond a few degrees, unlike photographic lenses.

[0033] 33. Indeed, a wide-field optic allows us to deduce information about the trajectory of the moving object; however, a wide field is more susceptible to being affected by stray light sources. Furthermore, it remains very difficult to design wide-field telescopes without encountering design problems, complex specific optical schemes, and enormous construction costs. The presence of a large focal plane also leads to numerous aberrations. When an electronic detector is coupled to a wide-field optic, it must be very large; the sizes and number of pixels can be very high, and the manufacturing and design costs are significant, making operation difficult. State of the art

[0034] 34. There are telescope systems that allow for very large fields of view by coupling a matrix of large-field telescopes forming a single field.

[0035] 35. French patent FR3018612 [WO2015136102] describes another known solution for detecting a moving object in space, characterized in that it comprises: a generation of a plurality of fields of view (Zkp) by means of a first set of telescopes (T), each telescope defining a detection telescope, the set of fields (FOV) of each telescope (Ty) having a spatial distribution in at least one plane of space inscribed in an open geometric shape (CC), said open geometric shape (CC) defining a wide detection field; a detection of at least one trace of a mobile (Mi) in the field (FOVy) of at least one telescope (Ty) by an electronic detector coupled to each telescope (Ty), the integration time of the electronic detector being defined to obtain a spreading of the trace over several pixels of the electronic detector for a given maximum speed (VM) of a mobile and a minimum altitude of its orbit; a deduction of a trajectory (TJSAT) of the mobile (Mi) in the image plane of said telescope (Ty).

[0036] 36. The company ASTRIUM proposes another solution described in patent document US2013 / 264463.This patent describes an optical system for a space surveillance system characterized in that it comprises an array of NxP telescopes each with a field of view greater than or equal to 5° and preferably greater than or equal to 10°, said telescopes being coupled to NxP image sensors of sensitivity adapted to an integration time on the order of 10 to 100 milliseconds, the telescopes being mounted on one or more motorized mounts, the telescopes being servo-controlled together and grouped so as to operate simultaneously to provide a wide field and in that the speed of movement of the mounts of the telescopes is such that each object crossing the scanned area is detected at least three times so as to obtain at least 3 dated position measurements distributed over the arc of transit of the object in the sky area, the exposure time or integration time being defined to obtain a spreading of the signal over several pixels.

[0037] 37. One drawback of this solution is the cost of such a system, which requires numerous very wide-field telescopes (several thousand). One solution is to reduce the number of telescopes and combine a motorized tracking system with wider-field telescopes having a field of view of at least 5°, and in practice 14° in the example cited in the patent document (10° x 10° on the square detector).

[0038] 38. US patent 7105791 describes a system for using an image of the Sun to detect objects traveling in the Earth's atmosphere. The system includes a receiver for collecting incident sunlight (solar energy) and a light-sensitive device that produces a signal in response to light exposure. A signal processor, coupled to the photosensitive device, detects the collected incident sunlight and is programmed to provide a corresponding output signal to detect a shadow moving across the photosensitive device.

[0039] 39. Patent EP1167997B1 proposes another solution for measuring space pollution, intended to be carried on a satellite, comprising: at least one laser illuminator capable of emitting a laser beam into space; means for receiving the signal back-reflected by space debris; means for detecting space debris that crosses the laser beam, determining the angular position of said debris; means for localizing the distance of the detected debris, determining the distances of the detected debris from the satellite using the impulsive and / or modulated nature of the laser beam emission; means for classifying the localized debris, determining for each localized debris the product of its average albedo by its apparent surface. Disadvantages of prior art

[0040] 40. Prior art solutions require expensive equipment to ensure significant coverage of the celestial vault (several tens of degrees), with wide field-of-view telescopes or with a large number of ground-based or satellite-mounted telescopes covering a large field. Solution provided by the invention

[0041] 41. To address these drawbacks, it is essential to catalog all potentially hazardous debris and assign valid orbital parameters to describe their trajectories. Observed from a fixed point on Earth, objects in low Earth orbit are characterized by their rapid movement across the sky. Furthermore, at any given moment, several objects traverse the sky at various locations. Depending on its orbital parameters, each object crosses the local sky at more or less regular intervals, ranging from a few tens of minutes to several hours.

[0042] 42. To this end, the invention, in its most general sense, relates to a system for detecting the trajectory of moving objects, characterized in that it comprises a rotating platform in an azimuthal plane supporting a plurality of telescopes, each oriented at a height between 35° and 85°, each telescope having a field of view between 2 and 6 square degrees and comprising a sensor of N x M pixels, each with a width L. The system includes a computer for recording the time-stamped images provided by the sensor of each telescope and for calculating the trajectory of a celestial object based on the light trails of said celestial object on a first image I1 and on a second image I2.

[0043] 43. Advantageously, the said platform is animated by a "step-by-step" jumping rotation.

[0044] 44. Preferably, said platform supports either four telescopes spaced 90° apart, or 6 telescopes spaced 60° apart, or 8 telescopes spaced 45° apart.

[0045] 45. According to a particular embodiment, the telescopes rotate step by step at a rate of several degrees per second, depending on the angle between the telescopes, the field of view of the telescopes, and their elevation relative to the horizon. A detailed description of a non-limiting example of this embodiment is provided.

[0046] 46. ​​The present invention will be better understood upon reading the following description, concerning a non-limiting example of an embodiment illustrated by the accompanying drawings where: [ FIG. 1 ] There figure 1 represents a schematic perspective view of a rotating platform according to the invention. FIG. 2 ] There figure 2 represents a schematic cross-sectional view of a rotating platform according to the invention. FIG. 3 ] There figure 3represents a schematic view of two shots. FIG. 4 ] There figure 4 represents a schematic view of the kinematics of the platform carrying the rotating telescopes relative to the platform. FIG. 5 ] There figure 5 represents a schematic view of a telescope's movement cycle. FIG. 6 ] There figure 6 represents a schematic view of the geographical distribution of telescope arrays. FIG. 7 ] There figure 7 represents a schematic view of a predicted collision and highlights the uncertainties in the positions using ellipsoids, allowing the collision probabilities to be established. FIG. 8 ] There figure 8 represents the functional architecture of image processing. FIG. 9 ] There figure 9 represents the functional architecture of processing block 1 for local data preprocessing. FIG. 10 ] There Figure 10represents the functional architecture of object classification processing block 2. FIG. 11 ] There figure 11 represents the functional architecture of blocks 4 to 7. Principle of the invention

[0047] 47. The invention aims to provide a means of optically detecting celestial objects within a very wide solid angle of approximately 1 steradian (between 30 and 120 degrees in a plane), with sufficient resolution to detect an object a few centimeters in cross-section from low Earth orbits to geostationary orbits. (This remark is to counter the argument that optics cannot be used in low Earth orbits due to the speed at which satellites pass.) No observation facility currently possesses optical characteristics that meet such specifications; only a network of a large number of telescopes would make it possible to achieve this performance, but this would be prohibitively expensive.

[0048] 48. The invention is based on: A first postulate that a celestial object will necessarily pass through an annular corona centered around an axis, which can be zenithal, polar, or chosen. A second postulate that it is possible to cover such an annular corona with an array of N telescopes with a limited field of view of 360 / N. and a third postulate that it is possible to emulate such a network with a smaller number of telescopes that are rotated at a sufficient speed so that the object is seen during its movement at least three times.

[0049] 49. The system according to the invention allows The detection of small objects, in massive quantities, in low Earth orbit that are too fast to be detected by conventional optical methods, unlike the use of radar technology in general. Detection is achieved by observing the traces left by the object in the images (standby / passive mode – as opposed to "tracking" mode): timestamping of the two ends of the line (opening and closing timestamped), and the length of the trace. Through astrometric reduction and the use of exposure timers (sub-millisecond), it is possible to determine an initial trajectory, the accuracy of which increases with each observation. The collection of all reconstructed trajectories allows for the creation and maintenance of an object catalog over time.

[0050] 50. The technical advantage resulting from the invention is to allow a limitation of the number of stations in the world; with a large field (3 to 6°) and a limited number of telescopes per station, while preserving a very high detection sensitivity to capture small objects and an ability to create and maintain a catalogue of objects on all orbits, but particularly in low orbit. Main features of the invention

[0051] 51. The system is based on the continuous observation of an observation ring at a specific elevation, in order to capture all objects that pass through this ring (unlike observing a section of the sky). The system takes advantage of the transit time of an object across the telescope's field of view (at the time of image acquisition) to rotate the telescopes around a vertical axis (pointing towards the zenith) without missing object detection. 52. In the case of four observation stations (each comprising four telescopes) and for an object transit time within a 5° field of view, the stations are not all located at the same latitude to compensate for seasonal variations and allow for the capture of objects at all latitudes. They are also distributed along longitudes to ensure observation is always possible during the night. 53.An object at 400km with a relative average speed of 0.5° / s takes 10s to traverse a 5° field. During these 10s, camera N+1 must have replaced camera N before the object leaves the field. 54. The relationships between: V: rotation speed of the object relative to the observer's position at a given altitude and elevation; C: vertical and horizontal field of view of the camera; N: number of telescopes; T: time taken by the object to traverse the field; CP: number of horizontal rotation steps of C degrees per second; E: elevation of the object relative to the horizontal plane at the observer's feet. Are

[0052] T × V = C P x C = 360, cos E / N / T in the example. P x 5° = 360 / 4 / 10 => P = 1.5 steps per second

[0053] 55. The total number of steps with N telescopes is: for E = 0° => Ptot = 360 / C; E= 40° => Ptot = 360 x cos 40 / C

[0054] With P = Ptot / C / T. 56. The choice of elevation is a compromise: The lower the object is on the horizon, the slower its relative rotation speed will be, but the farther away the object will appear. The closer the object is to the zenith, the closer it will appear, but the faster its rotation speed will be. The exposure time for each image is linked to the object's instantaneous speed as a function of its altitude. This is the time the object spends in a pixel, multiplied by at least 50, so that the tracks are at least 50 pixels long for proper automatic detection. 57. For this reason, the telescope's rotation speed is slaved to the elevation: The lower the elevation, the slower the object. The closer it is to the zenith, the faster it is. An alternative using rotating platforms

[0055] 58. These postulates lead to a solution consisting of a platform supporting several rotating telescopes—four in the example described by way of non-limiting example—for a Space Situational Awareness (SSA) application. This application is used for monitoring near-Earth objects, detecting natural objects such as asteroids and comets that could impact Earth, and for space surveillance, including tracking artificial satellites and active and inactive space debris, and estimating trajectories and collision risks. The objective of the invention is to optimize the number of telescopes required for surveillance activities in order to achieve comprehensive monitoring of orbiting objects at a fraction of the cost of current installations.

[0056] 59. This method makes it possible to compensate for the limited number of telescopes compared to the previous solution, by a scan obtained by the synchronized rotation of optical measuring devices each pointing in opposite directions at an angle of 360° / N telescopes.

[0057] 60. The observation ring detection system allows each object to be detected at two widely separated locations in the sky. This, combined with double measurement as the object passes through the ring's field of view, results in a large number of observed positions with very high accuracy. These four positions in space and time allow for a very good immediate approximation of the object's trajectory.

[0058] 61. This implies that the computing system must be able to recognize that the object measured once on the observation ring is indeed the same as the one measured at a given position a second time on the observation ring. The two measurements taken during the first crossing of the ring allow for a prediction of the second crossing. Therefore, the algorithm must verify the detected traces in the vicinity of the predicted time and position in order to establish the correspondences with an optimal degree of confidence. Physical description of the platform

[0059] 62. The figures 1 and 2 schematically represent a platform according to the invention. This platform (10), placed in an azimuthal plane, carries four telescopes (1 to 4) spaced 90° apart. Each telescope is driven in rotation about a zenithal axis with a jerky, step-by-step motion, which will be detailed below.

[0060] 63. The platform (10) supporting the telescopes (1 to 4) is fixed. Optionally, it can be animated by a constant rotation, preferably a jumping rotation, in a manner synchronized with the rotation of the telescopes (1 to 4).

[0061] 64. Alternatively, the telescopes can be animated with a constant rotational movement, the link between each telescope (1 to 4) and the platform (10) being ensured by a mechanism oscillating in a tangential direction around a median position.

[0062] 65. The telescopes have a height of 66° and a field of view of 4 square degrees. Each telescope has a sensor of NM pixels.

[0063] 66. Space objects are detected as streaks in astrophotography images that are processed for RA-DEC conversion and orbit determination. The ability to detect space objects depends on the time the object spends on a pixel, particularly in the case of low-Earth orbit objects at altitudes of around 2000 km, where angular velocity is high. Therefore, increasing the exposure time does not improve detection, as is the case in conventional celestial photography.

[0064] 67. By way of example, telescopes (1 to 4) are central obstruction reflecting telescopes with a large diameter primary mirror (e.g., astrographs with the sensor located in place of the secondary mirror), exhibiting the following characteristics Focal length: 790 millimeters Aperture: 356 millimeters Obstruction: 0.44% Resolution: 0.39 arcseconds Sensor: 4000x4000 pixels with a width of 15 microns.

[0065] 68. For an object with a diameter of 5 centimeters located at 1000 km, the parameters of a space object residing at a distance of 1000 km altitude are given.

[0066] 69. The time on a pixel is 6.5 ms for a 15-micron pixel. Therefore, if we consider a minimum trail length of 50 pixels (for correct trail detection), the minimum exposure time would be 300 ms. The same calculation at 400 km gives an exposure time of 150 ms.

[0067] 70. It is necessary to obtain at least three observations of the same object to perform a correct initial determination of the orbit, with the following consequences: It is not necessary to capture an entire section of the sky, but rather to capture the same object in different locations. This allows for capturing only a "slice" of the sky; the important thing is detecting the object as it crosses the slice, as described in the... figure 2 .

[0068] 71. An object in orbit needs a few minutes to cross the sky and a few seconds to pass through the telescope's field of view. Since one or two streaks (31, 41) are sufficient in A and B, continuous capture is not necessary. What is important is capturing the same object twice, but this can be done with an interval of several seconds, in two images (30, 40) as schematically illustrated by the figure 2 .

[0069] 72. The rotation of the telescope exploits the time interval between two images (30, 40) to capture the object a minimum of 2 times on 2 different images. Image acquisition

[0070] 73. With four telescopes per platform and to cover 90° in azimuth at an elevation of 66°, a field of view of 4° is possible. This represents 23 s between the first and last image (4° / 0.17) for objects at a distance of 2000 km. On lower orbits, this time interval could be shorter (13 s at 400 km).

[0071] 74. Therefore, if two telescopes (1, 2) are pointed outwards at an angle of 90°, there are 23 seconds remaining for the second camera (2) to assume the position of the first camera (1). In this case, the upper left band is captured by camera (1), and the lower right band by camera (2), as shown in the diagram. figure 4 Thus, a single telescope is sufficient to cover 90°.

[0072] 75. The figure 4schematically illustrate the situation where four pivoting telescopes (1 to 4) are supported by a platform (10), each on an ALT-AZ mount, and each oriented at 90° to each other. Assuming that every second the telescopes rotate 11.25° (1 / 8 of 90°), after 8 seconds, camera (2) can capture the streak that was captured by camera (1). In this example, if the time required for camera (2) to take the place of camera (1) is less than 23 seconds, no object is left undetected, as illustrated by the Fig. 4 .

[0073] 76. In this example, the movement sequence, over one second, is as follows: Rotation of 11.25°. Stop. Triggering of each camera is triggered for 300ms. Rotation of 11.25°. Etc...

[0074] 77. The platform (10) is optionally animated by a jumping rotation, exhibiting a speed profile similar to that of the hand of a jumping hour clock, with an angular step of 11.25° comprising an alternation of rapid movement of 11.25°, stop during the image acquisition time, about 300 ms, redeployment of one angular step, and so on.

[0075] 78. Furthermore, sidereal tracking is constantly in operation, to avoid the telescope stabilization time at the start of tracking. Data format delivered by each telescope (1 to 4)

[0076] 79. Each telescope is equipped with an NxM pixel sensor that provides digital images corresponding to the exposure time. Each image contains a background of fixed stars as well as traces formed by streaks. These streaks correspond to the movement of moving objects during the time between the beginning and end of image acquisition. The acquisition time is determined so that a streak covers a median value of 50 pixels.

[0077] 80. Each image allows data to be extracted in the form of time-stamped coordinates with a resolution of at least one millisecond from the beginning and end of the streak.

[0078] 81. These coordinates are determined by astrometric reduction from a catalog, for example SKY2000 or TYCHO-2, or USND-SA (trade names), using a matching algorithm. As an example, approximately twenty stars can be distinguished in a 3°x3° field with an exposure time of 300 ms, forming patterns that can be characterized by a matching algorithm with data from a star catalog.

[0079] 82. All the data collected for each image from each telescope are centralized on a server to allow the reconstruction of the trajectories of moving celestial objects whose tracks have been recorded in different images in a time-stamped format. Estimating the orbital trajectories for each object makes it possible to determine whether or not each object belongs to the orbital trajectory, thus building a catalog of orbital trajectories of the observed moving objects. Application example

[0080] 83. The table below gives the values ​​of average rotation speed of space objects in the observer's frame of reference, the time spent by the object in the field angle of the camera pixel and the exposure time required to produce a trace of 50 pixels. (RSO = resident space object).

[0081] 85. In the example at an altitude of 250 km, the average rotation speed of the object in the station's frame of reference is 0.38° / s. It takes 10.5 s to traverse the entire 4° field of view. This is the maximum time for CAM2 to capture an image of the same object as CAM1 while it remains within the field of view.

[0082] 86. The table below shows the different rotation speeds of the telescope, depending on the elevation of the observation ring.

[0083] 88. The indicated rotation speed values ​​are average values. For example, with an elevation of 66.4°, a 4° rotation step must be completed in 1.17 seconds. During this time, the telescope mount must accelerate to its nominal rotation speed for a period calculated based on the motor's capabilities, then decelerate. The camera must be triggered for a minimum of 115 milliseconds before the rotation process can begin again.

[0084] 89. The figure 5This illustrates the simplified kinematics of the operation. The rotation cycle lasts 1.17 seconds. During this cycle, the telescope alternates between an angular displacement phase (50) and a stationary phase (60) for imaging. The displacement phase (50) includes an acceleration step (51) until a steady rotation speed is reached (52), followed by a deceleration step (53) and a brief stabilization period, seemingly without movement, before the imaging phase (60) in the stationary position. Station characteristics

[0085] 90. The system comprises a minimum of four stations, with locations meeting several criteria: 91. The sky must be of very good quality from an astronomical point of view, in areas far from any light pollution. This is reflected in a sky magnitude (the same magnitude that defines the brightness of stars) and must be greater than magnitude 19 / °^2.

[0086] 92. The site should preferably be at altitude in order to suffer as little as possible from atmospheric turbulence.

[0087] 93. Meteorology must allow a minimum observation potential of 75% over the year.

[0088] 94. There should preferably be as many stations in the Northern Hemisphere as in the Southern Hemisphere, so as to obtain a good distribution of measurements to improve the quality of the calculated orbits.

[0089] 95. The position of the stations should allow as far as possible a complete view of the geostationary arc in order to ensure its complete monitoring.

[0090] 96. The number and position of the stations determines the possibility of continuous acquisition, which is a function of the shift in nighttime coverage.

[0091] 97. The figure 6This shows an example of station locations for a 6-station network (MTOS), each station comprising four rotating telescopes. Potential countries meeting these criteria include Morocco, the Canary Islands, Chile, Australia, Namibia, New Mexico, and Japan.

[0092] 98. To measure the performance of the stations, a geometric and photonic simulator was developed. It allows the simulation of the movements of a population of 1,000,000 objects larger than 1 cm around the globe and highlights:

[0093] 99. Objects passing through the observation rings of each station.

[0094] 100. Object detectability taking as a reference a signal-to-noise ratio of 5 at the end of the acquisition chain on the digital sensor.

[0095] 101. Depending on the characteristics of the telescopes and cameras chosen, in addition to the location of the stations, it is possible to estimate that the stations will allow the creation of a catalogue of several tens of thousands of space objects in Earth orbit, all orbits combined. Data processing.

[0096] 102. Astrophotographic images of space objects are part of a complete processing chain that determines the orbital trajectories of the detected objects. These trajectories are then propagated, an operation that consists of determining the object's position in the future (approximately ten days). This allows for the calculation of the closest approaches between the most critical objects and for the very precise calculation of the parameters of the probable collision (date and time of collision, distance between objects, probability of collision, spatial distribution of this probability).

[0097] 103. The figure 7 is an illustration of a predicted collision and the highlighting by ellipsoids of the uncertainties on the positions, allowing the establishment of the probabilities of collision. Data processing

[0098] 104. The figure 8 represents a schematic view of the functional architecture. The blocks operate in connection with each other, forming a coherent computational entity loop, where: Block 1 concerns local processing, at the station level, for data acquisition, image preprocessing, trajectory detection, astronomical calculations, and initial orbit determination. Block 2 concerns processing for detecting differences between real and virtual images and analyzing these differences, classifying objects, optimizing queues, and calculating the state vector database. Block 3 concerns the synchronization of local databases, the centralization of calculations, and digital data security. Block 4 performs orbital projection and numerical integrations. Block 5 concerns the calculation of distances between objects and the construction of a distance matrix. Block 6 concerns the calculation of risks, probabilities,Distances and timing of collisions. Block 7 concerns the probabilistic calculation of collision risks and densities, as well as the calculation of navigation, avoidance, and maneuvering instruction data. Block 8 concerns the programming of processing based on observation priorities.

[0099] 105. Block 1 concerns the processing of data from camera systems and telescopes for each of the local stations. This processing is performed in dedicated, local computing units. It includes image enhancement, trail detection, astrometric reduction, and initial orbit determination.

[0100] 106. The algorithms used for trail detection are known and can be improved using a supervised machine learning algorithm. The data flow is approximately 1 image / s, potentially for 24 rotating telescope systems across 6 MTOS (Multi-Telescope Observation Stations). The typical image size is 32 Mbytes (monochromatic, 16-bit encoded, 16-megapixel images). No buffer is planned to reduce processing time, and trail detection eliminates irrelevant images, thus reducing the required storage capacity. It is important to note that the output of this first processing block is text data of very small size. Indeed, transmitting image data is not feasible due to network bandwidth limitations in remote locations and over long distances.

[0101] 107. The figure 9This represents the functional architecture of object classification processing block 2. Block 2's function is to continuously update the catalog of spatial object trajectories. Therefore, the databases at local observation sites and the centralized databases must be constantly synchronized.

[0102] 108. In order to optimize the management of the spatial object queue, the following steps are implemented in this block 2: Periodic calculation of a theoretical image and comparison with the actual image, for example, for each second of acquisition. Identification and analysis of the differences between the theoretical and actual images. All uncertainties related to data acquisition appear at this stage: instrumental error, atmospheric disturbances, space objects with varying illumination due to rotation or tumbling, satellites missing following a maneuver, satellites reappearing following a maneuver, new space objects caused by a relaunch, partial separation of an existing satellite or collision or explosion, new space objects discovered through improved detection.

[0103] 109. A very promising methodology for improving the performance of the queuing process remains increasing the correlation between objects detected at the position and time of different objects with image sets from all MTOS. Ideally, the algorithm should learn to detect, with a certain level of confidence, the relationship between two objects separated in space and time. The result would be stored in a temporary list, pending further measurement to confirm or reject the correlation if no correlation can be established after a certain time.

[0104] 110. Difference analysis between synthetic and real images is not a simple task. Each trajectory must be sent to a subcategory of the main database. The simplest case is when the object is directly identified as corresponding to another object in the database. This means that the position in the image remains within a tolerance relative to the predicted position. A second layer of analysis is required for all other cases. If this is not performed correctly, the number of objects in the queue increases dramatically, to the point where the data becomes unusable. This part of the process allows each object in the trajectory data stream to be classified, with the smallest possible buffer size in the queue, in near real time.Artificial intelligence could be a solution to address the problem and improve performance over time and with experience. At the end of this block, all databases are identical across different sites and include all object state vectors containing: 3 values ​​for position, 3 values ​​for velocity, and 36 values ​​for the covariance of each object in position and velocity. Other elements can be added to the state vector, depending on the calculations required (such as the mass-to-surface-area ratio in the case of rotation or tumbling, or photometric characteristics of the object).

[0105] 111. Block 3 concerns real-time synchronization between the various local databases of the MTOS sites and the central computing center. The quality of the synthetic image depends on the completion of the database at any given time for each station and each local system. Because the MTOS stations are geographically dispersed, the network hardware must be sufficiently robust and reliable to achieve synchronization as quickly as possible. The network and links must be physically and cyber-secured to ensure the best possible protection of sensitive data. It should be noted that the data circulating between the stations to achieve this synchronization is text data.

[0106] 112. Block 4 concerns the calculation of the database state vectors and the calculation of orbital propagations with numerical integration at 15-second intervals over a 5-day period. Between 50,000 and 100,000 elementary calculations are required to numerically integrate one state vector with another, potentially for several thousand or hundreds of thousands of objects. For an object entering the database, an average of 75,000 elementary calculations are required, and with a 5-day propagation period at a 10-second integration interval, this represents 3 x 10⁹ elementary calculations. Based on 200,000 measurements per day, the average frequency would be more than 20 incoming data values ​​per second, which represents 6 x 10¹⁰ calculations per second.

[0107] 113. Block 5 concerns the calculation of distances between all objects for each integration time. As a basic assumption, we can consider that half of the objects cannot collide with each other. This would represent 100,000 x 100,000 matrices containing 2.5 x 10⁹ values, with a total of 8,640 steps over 5 days, which equates to more than 40,000,000 matrices to maintain.

[0108] 114. Block 6 concerns the calculation of collisions using the differences between a time T+1 and a time T for a given pair of objects. The time and distance of collisions can be calculated with a related probability, resulting in another set of 40,000,000 matrices.

[0109] 115. Block 7 concerns the intelligence layer fila e, which deals with collision probabilities and densities, global navigation management and automatic generation of instructions to satellite operators.

[0110] 116. Block 8 uses data from block 7 and possibly block 6 to establish observation priorities and the resulting instrument programming, particularly when congestion becomes significant or special measures need to be carried out (laser rangefinding, for example).

Claims

1. A system for detecting the path of moving celestial objects using a system formed by a plurality of telescopes (1 to 4), each rotating about its zenith axis, each oriented at an elevation between 35 and 85°, each of said telescopes having a field of view ranging from 2 to 6 square degrees and comprising a sensor with NxM pixels, each with a width L, the system also comprising a central computer to record the supplied time-stamped data corresponding to each image from each sensor of each of said telescopes (1 to 4), wherein said data comprise, for each image, timestamped coordinates with an accuracy of at least one millisecond for the start and end of each trace recorded during the exposure time, said coordinates for the start and end of each trace being determined by astrometric reduction by matching the fixed-star pattern of each image with the data from a star catalogue, said central computer executing processing operations to estimate the orbital trajectories of each moving object in order to determine, for each of said objects, whether or not it belongs to an orbital path, as well as processing operations to record a catalogue of the orbital parameters of said moving objects.

2. The system for detecting the path of moving objects according to claim 1, wherein said telescopes (1 to 4) are driven in a jumping rotation.

3. The system for detecting the path of moving objects according to claim 1 or 2, wherein the number of rotation steps per second of said telescopes (1 to 4) is determined by the equation P x C = 360 . Cos E / N / T with: • V: rotational speed of the object relative to the position of the observer at a given altitude and elevation C: vertical and horizontal field of the camera • N: Number of telescopes • T: time lapsed while the object crosses the field C • P: number of steps required to rotate by C horizontal degrees per second • E: Object elevation with respect to the horizontal plane at the foot of the observer • C = T × V .

4. The system for detecting the path of moving objects according to claim 1, wherein the system supports four telescopes (1 to 4), each spaced 90° apart.

5. The system for detecting the path of moving objects according to claim 1, wherein the system comprises a platform (10) supporting four telescopes (1 to 4) whose observation axes spaced 90° apart.

6. The system for detecting the path of moving objects according to the preceding claim, wherein said platform (10) is driven in a jumping rotational movement with steps of 11.25°.