• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    A method of tracking a transfer orbit or a phase of orbiting a spacecraft comprising the steps of: a) tracking at least one GNSS signal (Sgnss) and using it to determine at least one pseudo distance between said spacecraft and one or more GNSS satellites transmitting said or each said signal; b) using an estimation model (ME) to jointly estimate a set (ξ) of state parameters of said space vehicle comprising a plurality of position parameters, a plurality of speed parameters and at least one error parameter of thrust characterizing a difference between a real thrust force of said spacecraft and a nominal thrust force by taking said one or more pseudo-distances as input data of said estimation model. Apparatus for implementing such a method
    公开号:FR3018121A1
    申请号:FR1400508
    申请日:2014-02-28
    公开日:2015-09-04
    发明作者:Thibaud Calmettes；Quillere Fabien Rozo；Damien Serant
    申请人:Thales SA；
    IPC主号:
    专利说明:

    [0001] METHOD FOR TRACKING A TRANSFER ORBIT OR A PHASE FOR ORKING A SPATIAL VEHICLE, IN PARTICULAR AN ELECTRICAL PROPULSION, AND APPARATUS FOR IMPLEMENTING SUCH A METHOD The invention relates to a method for orbital tracking of a spacecraft by means of a Global Navigation Satellite System (GNSS) of the GPS, Galileo or GLONASS type, as well as an apparatus for the implementation of such a method. The invention applies more particularly to tracking a transfer orbit - particularly to a geostationary orbit - or a phase of orbiting a spacecraft, especially when the latter uses an electric propulsion. Satellite navigation systems have been designed to allow the tracking and tracking of ground user trajectories, or at most aircraft operating at low altitude. However, it is also known to use them for the tracking of orbits of space vehicles, mainly in low Earth orbit (LEO), much more rarely in the case of intermediate orbits (MEO, "Medium Earth Orbit") or geostationary (GEO). However, these systems are not used to date for the monitoring of the transfer phases or in orbit, because the constraints of use are strong: the GNSS signal is generally weak because these orbits are outside the main lobes of the antennas transmitters of GNSS satellites (which are logically pointed to the earth); there are significant periods of non-availability; the activation of the propulsion leads to important location errors and prevents the use of the simplifying assumption that the vehicle follows a Keplerian orbit, etc. These constraints are even more important in the case of space vehicles with electric propulsion (ion emission, for example), in which the thrust is continuous over the duration of the transfer phase. For these reasons, the orbital positioning being transferred is achieved through distance measurements on a TMTC communication channel between the spacecraft and earth stations. This solution has many disadvantages: need to use several ground stations to have sufficient geometric diversity, which entails a significant cost; inaccuracy of location; delay between measurements which induces strong drifts; low autonomy in case of degradation of the communication link. The invention aims to remedy the aforementioned drawbacks of the prior art. More specifically, it aims to make possible GNSS tracking transfer orbits and orbiting phases of space vehicles, including continuous thrust and more particularly electric propulsion. According to the invention, such a goal is achieved by virtue of a tight coupling between a GNSS processing and an estimation model of a spacecraft state vector taking into account the propulsion of the latter. Thus, an object of the invention is a method of tracking a transfer orbit or a phase of orbiting a spacecraft comprising the following steps: a) tracking at least one GNSS signal and using it to determine at least one pseudo-distance between said spacecraft and one or more GNSS satellites transmitting said or each said signal; b) using an estimation model to jointly estimate a set of state parameters of said space vehicle comprising a plurality of position parameters, a plurality of speed parameters, at least one clock parameter, and at least one parameter thrust error characterizing a difference between a real thrust force of said spacecraft and a nominal thrust force, by taking said one or more pseudoranges as input data of said estimation model. According to particular embodiments of such a method: Said thrust may be a continuous thrust, in particular of electric type. Said step a) can comprise the ground assisted recovery of the GNSS satellite navigation message or of each navigation message of each GNSS satellite. Said step a) may comprise the tracking of at least one pilot GNSS signal, not carrying a navigation message, said tracking being carried out by coherent correlation of said or each said pilot GNSS signal for a duration greater than that of a symbol. Alternatively, said step a) may comprise an operation of erasing said or each said navigation message 10 recovered by assistance from the ground to enable said coherent correlation. Said set of state parameters may comprise at least a first thrust error parameter characterizing a difference between an amplitude of said actual thrust force and said nominal thrust force, and a second thrust error parameter; characterizing a difference between a direction of said actual thrust force and said nominal thrust force. More particularly, said set of state parameters may comprise at least two push error parameters characterizing both said deviation between a direction of said actual thrust force and said nominal thrust force, and an error of estimate of the attitude of said space vehicle. More particularly, the method may also comprise the following step: c) applying, for a determined duration, a thrust nominally directed in a direction connecting said space vehicle to a GNSS satellite transmitting a GNSS signal acquired during said step a ); whereby the application of said estimation model allows an improvement in the estimate of said or at least one of said thrust error parameters. Said set of state parameters may also comprise at least one clock error parameter characterizing an offset or a drift of a clock of a GNSS receiver embedded on said space vehicle and used to implement said step at). Said estimation model can be a Kalman filter or an extended Kalman filter characterized by a matrix or transition function depending on the time and the position of said space vehicle so as to take into account both gravitational forces acting on said spacecraft and said thrust force. The method may also comprise the following step: d) using pseudo-distances measured in the past, as well as the corresponding estimates of said state parameters, to update a covariance matrix of said estimation model. The method may also include the following step: e) using said at least one said acquired GNSS signal, a knowledge of the ephemerides of the GNSS satellite having emitted it and the position of the estimated space vehicle by means of said estimation model for estimate a radiation pattern of an antenna transmitting said GNSS satellite. The method may also comprise the following step: f) using said or at least one said estimate of a radiation pattern during the implementation of said step a). The method may also comprise the following step: g) using the knowledge of said estimation model and said state parameters to determine future corrections of the orbit and / or the thrust of said space vehicle. Another object of the invention is an apparatus for implementing such a method comprising: a GNSS receiver configured to track at least one GNSS signal; A data processing device receiving as input a GNSS signal tracked by said receiver; said processing device being configured or programmed to: use said or each said acquired GNSS signal to determine at least one pseudo-distance between said receiver and one or more GNSS satellites transmitting said or each said signal; - applying an estimation model to jointly estimate a set of state parameters comprising a plurality of position parameters, a plurality of speed parameters, at least one clock parameter, and at least one thrust error parameter Characterizing a difference between an actual thrust force of said spacecraft and a nominal thrust force, by taking said one or more pseudoranges as input to said estimation model. Other features, details and advantages of the invention will become apparent on reading the description given with reference to the accompanying drawings given by way of example, in which: FIG. 1 shows a space vehicle whose transfer orbit is followed by a method according to the invention as well as a GNSS satellite used for the implementation of this method; FIG. 2 illustrates the relationship existing between a local orbital reference and a satellite reference; FIG. 3 illustrates the notion of attitude error; Figure 4 is a diagram of a method according to an embodiment of the invention; and Figure 5 is a block diagram of an apparatus according to one embodiment of the invention. Figure 1 illustrates a configuration in which the present invention can be implemented advantageously. In this figure the reference T corresponds to the Earth and OT to its center serving as origin to the reference ECI ("Earth-Centered Inertial", that is to say "Inertial Centered on the Earth"), whose axes , mutually orthogonal, are XECI, YECI and ZEci. A low-Earth orbit SAT satellite, belonging to a GNSS constellation, transmits a positioning signal by means of an antenna having a DR radiation pattern whose main lobe is directed towards the earth. The spacecraft VS moves in an OVS transfer orbit at an altitude higher than that of the satellite SAT. The OVS orbit is not Keplerian, because the propulsion - for example electric - of the spacecraft is activated; in addition, the navigation signal is received with a low signal-to-noise ratio since only one secondary lobe of the radiation pattern DR can be picked up by a receiving antenna of the spacecraft VS. The figure illustrates a single GNSS satellite SAT, but it should be understood that in general several of them can be in visibility conditions from the space vehicle 10 VS. Figure 2 shows schematically the spacecraft VS and two reference centers centered on its center of mass cm, which moves along the OVS path. The mutually orthogonal axes XROL, YROL and ZROL form the "orbital reference"; conventionally one chooses XROL oriented 15 along the speed axis and ZROL such that the center of the earth OT is in the plane (XROL, cm, ZROL) - The mutually orthogonal axes XRS, YRS and ZRS form the "satellite reference", linked physically to the VS space vehicle. Conventionally one chooses XRS opposite the axis of the nozzle, so that the thrust is made in an XRS direction. The attitude of the space vehicle is the rotation between the two landmarks (XROL, YROL, ZROL) and (XRS, YRS, ZRS). Conventionally, it can be defined by three Euler angles or by quaternions. The attitude is generally estimated by methods such as the use of stellar sensors or taking into account the history of rotations of the attitude control wheels. This estimate inevitably has an error ("attitude error"), the consideration of which is an important aspect of the present invention. The attitude error can be defined by the error between the "true" XRS axis and the "estimated" XRS axis. This error is a rotation, usually defined by three angles; however, since it is generally small, it can be described by only two angles (azimuth offset, az offset): The relationship between the true satellite reference (XRS, YRS, ZRs) and the estimated satellite reference (5C-Rs, YRS 'ZRs) is illustrated in Figure 3. Figure 4 schematically illustrates an orbit tracking process, which will be described in detail below.
    [0002] The first step (a) of such a method is the acquisition of one or more GNSS signals and their use to determine one or more pseudo-distances. The GNSS signals are CDMA (Code Division Multiple Access) type spreading signals, consisting of a useful binary content (navigation message) multiplied by a spreading code. The spreading code is repeated with a periodicity equal to the duration of a symbol, with a positive or negative sign depending on the value of said symbol. In the context of the use of GNSS for navigation in high orbits we work at low signal-to-noise ratio and it therefore becomes particularly interesting to consider, among the signals, only the pilot paths, ie the tracks for which the useful binary content is fixed (symbols equal to zero or one). Thus, it is possible to integrate the signal coherently over a longer length than the duration of the useful symbol. With the exception of the Ll track on GPS, all the GPS and Galileo signals (as well as GLONASS and Compass) always consist of a pilot channel and a modulated data channel, so that this constraint of use pilot does not result in a frequency limitation (except for GPS Track L1). In return, the information carried by the navigation signal - and in particular the long-term ephemerides of the GNSS satellites - must be recovered by another means, typically, from a ground station. Alternatively, it is possible to use the navigation signal thus recovered to "remove" the binary symbols of a pilot channel, and transform the latter into a pilot channel ("data wipe-off").
    [0003] The use of a single pilot channel makes it possible to carry out an ambiguous pseudo-distance measurement at the code period. Indeed, all symbols being the same, we can date the beginning of a symbol but we can not say a priori which symbol it is. Depending on the band used, this ambiguity is between 1ms (300km in distance) and 4ms (1200km). Such distances are large enough to allow rapid ambiguity removal in ground applications, but can be fairly easily achieved in the context of orbital propagation, given the possibility of encountering long periods (several hours) without possibility of measurement. According to the invention, this ambiguity is solved by using a model for estimating the thrust of the space vehicle. This model can advantageously be integrated into an estimation model of a state vector including, in addition to parameters characterizing the thrust, the position and the speed of the spacecraft. In addition, the integration of the thrust in the estimated state vector makes it possible to improve the continuous knowledge of the position of the receiver, including in pure propagation phases where the GNSS satellites are not or little visible. It is therefore relevant to consider a so-called "tight" integration of GNSS measurements in the processing. Indeed, from the framework of a "loose" operation, the pseudo-distance measurements of the GNSS processing are carried out independently, and then transmitted to the navigation function which takes care of taking them into account in the trajectory tracking. . In "tight" operation, the navigation function provides GNSS processing with the estimated position of the receiver. By combining this estimated position of the receiver with the estimated position of the GNSS satellites, the GNSS processing can anticipate the reception frequencies and synchronization of the GNSS signals, which greatly facilitates its operation and which improves in particular its ability to operate at a low signal ratio. to noise. Advantageously, in one embodiment of the invention, the state vector (reference in FIG. 4) is estimated by Kalman filtering (block b) using an estimation model ME taking into account the times gravitational forces and thrust.
    [0004] For the record, the Kalman filtering comprises a first prediction phase, in which the state vector at time tk and its covariance matrix are estimated from the only data available at time tk_i, followed by a correction phase exploiting the measure of an observable at time tk. The equations of the Kalman filter are given by: - Prediction phase: 'kIk-1 = Fk-14-11k-1 + Bk-1uk-1 Pkik-1 = Fk-1P k-ilk-lFTc-1 + Qk - Correction phase: Yk = Zk + Hilik-1 Sk = HkPkik-11171; + Bk Kk = Pkik-111TcSk1 ekik = ekik_14- KIJ7k Pkik = (I-KkHk) Pki ik-1 Where: - The exponent "T" indicates the transposition operation, the exponent "-1" the inversion of 'a matrix ; - The symbols in bold denote matrices; - ekik-1 is the prior estimate of state; - Pkik-1 is the covariance of estimate a priori; - Fk-1 (resp Fk) is the state transition matrix at time tk-.1 (resp. Tk); - Bk_i is the command matrix; - Uk_1 is a control vector; - Zk is the result of measurement of the observable at time tic; - k is innovation; - Sk is the covariance matrix of the innovation; - Hk is the measurement matrix; - Kk is the optimal Kalman gain; - kik and P kik are, respectively, the state estimate and the posterior estimation covariance; - Qk and Rk are noises. In the following we will consider that the control vector u is identically zero. The state vector E comprises at least position and speed parameters, as well as one or more thrust error parameters, characterizing a difference between a real thrust force acting on the spacecraft and a nominal thrust force. . Advantageously, it may also include one or more parameters characterizing an error of the clock embedded in the GNSS receiver. As a specific, non-limiting example, we can have: = [x; y; z; vx; vy; vz; ho; hi; Go; of; where: - x, y, z are the coordinates of the center of mass of the spacecraft in an ECI frame (see Figure 1), and vx, vy, vz the corresponding components of the velocity; ho and h1 are, respectively, bias and drift of the receiver clock; - Go is an amplitude error of the thrust; de and da, defined above with reference to FIG. 3, of the parameters which characterize the attitude error, and therefore the angular error of the thrust, considering that the thrust has a fixed orientation with respect to the space vehicle body. It is easy to understand that other representations of the state vector are possible. For example, the position may be defined by a Keplerian elliptical representation; the drift of the clock can sometimes be neglected; we can introduce a thrust amplitude error drift, G1; the angular parameters of and da can be replaced by three parameters representing a roll, pitch and yaw error. From the state vector at a date tk1, the state vector 1 is calculated by propagating the force model (ik_1 at the date tk = tk-1 + At, for example by taking At = 30 seconds. means in practice that knowing the position (necessary for the evaluation of earth gravity, lunar and solar gravity forces), the thrust and the axis of the thrust (necessary for the evaluation of the thrust force), knows the acceleration exerted on the satellite at tk_i and can deduce the variation of all the parameters: - Position (klk-1) = Position (k) + Speed (k) - At - Speed (k / k- 1) = Velocity (K) + Acceleration (k) ot - ho (klk - 1) = ho (k) 25 - hi (klk - 1) = hi (k) - Go (klk - 1) = Go (k) - (klk-1) = of (k) - da (klk-1) = da (k) We can thus construct the variation matrix (Fk4 in the formulas above), in the form: F k-1 = 1 0 0 At 0 0 0 0 0 0 0 0 1 0 0 At 0 0 0 0 0 0 0 0 1 0 0 At 0 0 0 0 0 Al A2 A3 1 0 0 0 0 A4 A5 A6 A7 A8 A9 0 1 0 0 0 A10 A11 A12 A13 A14 A15 A16 A18 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 The values of Al to A18 are equal to At multiplied by the partial derivatives of the acceleration with respect to the position elements and the thrust elements. For example, Al = At Fr, where ï is the component of the acceleration along the x axis. In practice, if the interval At is too large to consider the position, speed and acceleration as constants, one can use Runge-Kutta type integration methods. Note that the equations of motion are nonlinear because gravitational forces depend on position. However, in the model above the equations are linearized and the non-linearity is reflected by the fact that the matrix F is not constant but changes from one propagation step to another. Likewise, the effect of the thrust is integrated in the matrix F, which makes it possible not to use a control vector u.
    [0005] It is interesting to note that the attitude of the satellite intervenes in the state transition matrix through thrust. This means that a variation of attitude results in a variation of position. Since the position is observable by the pseudo-distance measurement (see below), the attitude is observable. The fact that the attitude has an influence on the position is directly translated by the fact that the elements A5, A6, A11, Al2, A17 and A18 in the matrix F are non-zero. Thanks to the integration of the parameters of and da into the state vector, it is therefore possible to estimate continuously the estimation error on XRS and Rs and thus contribute to the attitude estimation of the satellite. In the Kalman filter model, the propagation error can be taken into account. The estimated error values must be introduced into the matrix Qk of the formulas above. For example, the acceleration model can be considered to have an error (standard deviation) of 1.10-7 m / s2 in each direction, for example because, for the sake of simplicity, the effect of lunar attraction. In this case we will commit on the propagated speed an error of 1.10-7-At m / s = 3.10-6 m / s taking At = 30s. Then, considering that there is no other error, we have: Qk = (3.10-6) 2. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15 the square being due to the fact that Qk is a matrix of variances and not of standard deviations. In the Kalman equations, a measure zk is used to calculate the innovation ÿk as the difference between the expected measure, function of klk-i, and the measure actually performed. In the case of GNSS, the measurement concerned is a pseudo-distance measurement. By considering the coordinates (XSAT, YSAT, ZSAT) of the GNSS satellite at the moment of the signal emission at the date tE (real date), and tR being the reception date determined by the receiver, the pseudo-distance PD is given by: PD = tR - tE = Ni (x xsAT) 2 + (Y YsAT) 2 (z ZSAT) 2 + hh being the error of the onboard clock (h = ho + hl-tk by taking the vector of previously defined state) and c the speed of light. The innovation ÿk is the difference between the pseudo-distance actually measured that expected at the time tk k = PD (xlc I k-1 XSAT) 2 f YSAT 2 f ZSATN 2 I + hk I k -1 C Only the parameters that have an impact on the pseudo-distance are observable directly, namely the clock error and the position. According to an advantageous aspect of the invention, it is possible to optimize the trajectory to make sure that the thrust has a strong influence - although indirect on the measurement of pseudo-distance. For simplicity, it is assumed that the GNSS SAT satellite is located at the center of the Earth, OT, and that the space vehicle VS is located at a position (0, 0, z) in the ETI frame; then, the matrix Hk (in fact, a line vector which, multiplied on the right by ilk_idonne a scalar) is: Hk = [0, 0, 1 / c, 0, 0, 1, 0, 0, 0, 0 ] Indeed pseudo-distance varies in proportion to the clock and in 20 1 / c with respect to the direction z, the other elements having no impact. A thrust along the XROL axis has an impact on the speed vx then, at the sufficient iteration, on the position x, but this has no impact on the pseudo-distance measurement, because the direction XROL is orthogonal to the axis linking the GNSS satellite to the spacecraft (parallel to ZROL). Therefore, in order to reduce the residual between the actual measurement and the estimated measurement, the Kalman filter will firstly correct the clock error parameter and the estimates of z and vz, i.e. which have an impact on the measurement, to the detriment of the other elements of the state vector, notably the thrust. If, on the other hand, the attitude of the spacecraft is modified so that the thrust is temporarily oriented along the ZROL axis, then the thrust has a strong influence on the pseudo-distance measurement, and the Kalman filter will correct the z-estimates. and vz, thrust errors (angular error expressed by de and da and amplitude error Go), besides the clock error parameter.
    [0006] In other words, by choosing the thrust axis, the observability of the thrust estimate of the state vector is improved. From a mathematical point of view, this results in the following way: - In the absence of contribution of the measurement on a variable, its covariance (the line which corresponds to it in the matrix P) increases more quickly during the the propagation step (which changes it from Pk-1 to Pkik-i) than it decreases during the measurement correction (which makes it go from Pkik_i to Pk). This reflects an increasing uncertainty for the filter on this value. On the other hand, if the contribution of the measurement is strong, there is a sharp decrease in the transition from Pkk 1 to Pk, in particular greater than the increase between Pk_i and Pkik-1; therefore, overall, the uncertainty is reduced between tk-1 and tk. The fact of choosing the thrust axis not only according to the desired direction with respect to the orbital transfer, but also according to what this allows as observability, and this especially in the cases where one would see the lines corresponding to the thrust increase strongly in the matrix P, is shown diagrammatically in FIG. 4 by the "GNSS measurement strategy" block (reference c). It is known from the theory of Kalman filtering that the state vector E can not be too complex or face inver- tability problems. This requires excessive simplification of space vehicle motion modeling. For example, the thrust amplitude error g is modeled by an offset, or even by an offset and a linear drift; but a more realistic model would be: g (t) = go + g1.sin (g24-Fg3) involving four parameters go, g1, g2 and g3.
    [0007] To overcome this difficulty one can resort to a "loopback on the past" (reference d in Figure 4). For example, consider the case where the state vector comprises a single parameter g (t) = GB constant modeling the thrust amplitude error whereas, as shown above, four parameters would be necessary. We have: - g (tk_i) = go + g1.sin (g2.tk-1 + g3) - g (tk) = go + g1.sin (g2.tk + g3) By putting tk = tk-1 + At: - g (tk) = go + gisin [g2. (tk_i-FAt) + g3] 15 which after sinus development and simplification to order 1 and considering small g2-At (in other words the sinusoidal variation period is much greater than At), or cos (g2- At) = 1 and sin (g2-At) = 0, gives: - g (tk) = g (tk-i) g1.g2.At.cos (g2.At + g3) 2 0 In other words: - Having estimated correctly Go = g (tk-i) = go + g1-sin (g2-tk-1 + g3) on the date tk-1, - and by propagating by g (tk) = GB, - we introduce an error of AG ° = g1.g2.At.cos (g2.tk_1 + g3) on propagation. Thus, if instead of working point by point with a Kalman filter, it is estimated at once, for example by a least squares estimation on the past day (24 hours), a much more complex state vector, integrating go, g1, g2, g3 as well as the equivalent parameters on the attitude and clock error, this least squares estimator has no problem of convergence since we consider many more measures than of parameters. If we measure all At = 30s, we have 24.3600 / 30 = 2880 measurements for, for example, 32 parameters (11 parameters defining the state vector and harmonic models - each introducing three additional parameters - for thrust amplitude error, the clock error, the two components of the angular error of thrust, the three acceleration components). After this least squares estimation, we can consider that these parameters are known for the following day. Thus, limiting itself to the case of the thrust amplitude error, in the following day, when applying the Kalman filter, we will add to the ninth row / ninth column of the covariance matrix Qk of the model. the value (AGo) 2 = (gi.g2.At.cos (g2.tk-i-Fg3)) 2. In other words, the "loopback on the past" consists in using previously determined pseudo-distances (for example, during a past day) as well as the corresponding estimates of the state vector, to update a matrix of covariance of the estimation model. This updated matrix will then be used - for example during a next day - for the implementation of Kalman filtering.
    [0008] When a GNSS measurement is carried out, the following elements are known: - Position of the GNSS satellite (by its ephemeris, necessary to transform the measurement of arrival time into useful pseudoranges); Space vehicle position (Kalman state vector); The power of the received GNSS signal, evaluated by conventional GNSS processing methods. These elements can advantageously be used to estimate the gain of the DG antenna pattern of the GNSS satellite transmitting antenna in the satellite-space vehicle direction. Indeed, the knowledge of the position of the GNSS satellite "SAT" and the spacecraft "VS" makes it possible to determine an elevation angle of VS seen by SAT and SAT seen by VS, as well as the distance SAT - VS. This in turn allows the link budget to be determined, as only the antenna gain of the GNSS satellite is unknown. This last parameter can be determined from a measurement of the power of the received GNSS signal.
    [0009] The estimation of this antenna gain (reference "e" in FIG. 4) can advantageously be used to optimize the GNSS tracking strategy (reference "f"). Indeed, the embedded GNSS receiver knows its own position (estimated by the Kalman filter), the position of the Earth and the position of all GNSS satellites. It can therefore calculate for each of these GNSS satellites the corresponding elevation angle and, from the antenna pattern estimated as described above, decide whether the GNSS satellite is usable or not, and therefore decide to achieve a re-acquisition and a pseudo-distance measurement on it or not.
    [0010] Advantageously, the knowledge of the estimation model and the state parameters of the vector E can be exploited to determine future corrections of the orbit and / or the thrust of said space vehicle. For example: Knowing the effective thrust curve, it will be best to distribute it over the rest of the positioning, for example by limiting out-of-perigee thrusts. Knowing the precise antenna pattern of each GNSS satellite, it will be possible to choose a priori the GNSS satellites tracked at each measurement point more effectively, and the associated measurement covariance will be more precisely determined. - Knowing the current position of the spacecraft, and comparing it to the position expected during the initial construction of the scenario, it will be known what additional course correction must be made to readjust as expected the final orbit; this particularly concerns inclination and nodal longitude. This step of "optimization on the future" is indicated in FIG. 4 by the reference "g".
    [0011] FIG. 5 illustrates a space vehicle VS carrying a device for implementing a method according to the invention. The spacecraft is equipped with a first ARgnss antenna to receive at least one GNSS signal sgnss and a second antenna ARite to receive, from the ground, at least one signal site representative of long-term ephemeris GNSS satellites and a star sensor SS to generate a signal satt representative of the attitude of the satellite. In other embodiments, a single antenna can acquire long-term GNSS and ephemeris signals and / or other means, such as control wheels, can be provided for attitude estimation. . The sgnss signal (s) picked up by the antenna ARgnss are supplied as input to a GNSS receiver (reference Rgnss) which, by a conventional processing, determines the corresponding pseudo-distances (possibly ambiguous). A DTD data processing device, typically a conveniently programmed onboard computer, receives as input spd signals generated by the Rgnss receiver, the sitt signal, the set signal, and a sp signal representative of a generated thrust. by an electric propulsion engine MPE. The DTD uses these input data, as well as an estimation model ME, to determine (among other things) the instantaneous position and velocity of the spacecraft. These data can be transmitted to a station on the ground, for example by means of a dedicated antenna (not shown) or the antenna ARIte. The invention has been described with reference to a particular embodiment, but many modifications are possible.
    [0012] For example, the Kalman filter can be replaced by another estimation model such as an extended Kalman filter, a genetic algorithm or a smaller iterative square.
    权利要求:
    Claims (15)
    [0001]
    REVENDICATIONS1. A method of tracking a transfer orbit (OVS) or a phase of orbiting a spacecraft having the following steps: a) tracking at least one GNSS signal (sgnss) and using it to determine at least one pseudo-distance between said spacecraft and one or more GNSS satellites (SAT) transmitting said or each said signal; b) using an estimation model (ME) to jointly estimate a set () of state parameters of said space vehicle comprising a plurality of position parameters, a plurality of speed parameters, at least one clock parameter, and at least one thrust error parameter characterizing a difference between an actual thrust force of said spacecraft and a nominal thrust force, by taking said one or more pseudo-distances as input to said estimation model .
    [0002]
    2. Method according to claim 1 wherein said thrust is a continuous thrust, in particular of the electric type. 20
    [0003]
    3. Method according to one of the preceding claims wherein said step a) comprises the ground assistance recovery of the GNSS navigation message or each navigation message of each GNSS satellite. 25
    [0004]
    The method according to claim 3 wherein said step a) comprises the tracking of at least one pilot GNSS signal, not carrying a navigation message, said tracking being performed by coherent correlation of said or each said pilot GNSS signal on a longer than that of a symbol.
    [0005]
    The method of claim 3 wherein said step a) comprises an operation of erasing said or each said navigation message retrieved by assistance from the ground to enable said coherent correlation.
    [0006]
    6. Method according to one of the preceding claims wherein said set of state parameters comprises at least a first thrust error parameter characterizing a difference between an amplitude of said actual thrust force and said nominal thrust force, and a second thrust error parameter characterizing a difference between a direction of said actual thrust force and said nominal thrust force.
    [0007]
    The method of claim 6 wherein said set of state parameters comprises at least two push error parameters characterizing both said deviation between a direction of said actual thrust force and said nominal thrust force, and an error in estimating the attitude of said space vehicle.
    [0008]
    8. The method of claim 7 also comprising the following step: c) apply, for a predetermined duration, a nominally directed thrust in a direction connecting said spacecraft to a GNSS satellite transmitting a GNSS signal acquired in said step a); whereby the application of said estimation model allows an improvement in the estimation of said or at least one of said thrust error parameters.
    [0009]
    9. The method according to one of the preceding claims, wherein said set of state parameters also comprises at least one clock error parameter, characterizing an offset or drift of a clock of a GNSS receiver embedded on said space vehicle and used to implement said step a).
    [0010]
    Method according to one of the preceding claims, wherein said estimation model is a Kalman filter or an extended Kalman filter characterized by a matrix or transition function depending on the time and the position of said space vehicle so as to take in account both gravitational forces acting on said spacecraft and said thrust force.
    [0011]
    11. Method according to one of the preceding claims also comprising the following step: d) using pseudo-distances measured in the past, as well as the corresponding estimates of said state parameters, to update a covariance matrix of said model of 'estimate.
    [0012]
    12. Method according to one of the preceding claims also comprising the following steps: e) using said at least one said GNSS signal acquired, a knowledge of the ephemeris of the GNSS satellite having emitted it and the position of the estimated space vehicle by means of said estimation model for estimating a radiation pattern (DR) of an antenna transmitting said GNSS satellite. 25
    [0013]
    13. The method of claim 12 also comprising the following step: f) use said or at least one said estimate of a radiation pattern during the implementation of said step a). 30
    [0014]
    The method according to one of the preceding claims also comprising the following step: g) using the knowledge of said estimation model and said state parameters to determine future corrections of the orbit and / or thrust of said vehicle spatial.
    [0015]
    Apparatus for implementing a method according to one of the preceding claims comprising: - a GNSS receiver (Rgnss) configured to track at least one GNSS signal; a data processing device (DTD) receiving at input a GNSS signal tracked by said receiver; said processing device being configured or programmed to: - use said one or each said acquired GNSS signal to determine at least one pseudo-distance between said receiver and one or more GNSS satellites transmitting said or each said signal; - applying an estimation model to jointly estimate a set of state parameters comprising a plurality of position parameters, a plurality of speed parameters, at least one clock parameter, and at least one thrust error parameter Characterizing a difference between an actual thrust force of said spacecraft and a nominal thrust force, by taking said one or more pseudoranges as input to said estimation model.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP2921923B1|2018-07-11|Method for tracking a transfer orbit or a phase of placing a space vehicle in orbit, in particular a vehicle with electric drive, and apparatus for implementing such a method
    EP3230766B1|2020-05-06|Method and system to validate geopositioning by satellite
    EP2245479B1|2016-11-30|Navigation system using phase measure hybridisation
    EP2410352A1|2012-01-25|Antenna device with synthetic opening for receiving signals of a system including a carrier and a means for determining the trajectory thereof
    EP3505968B1|2020-09-16|Method for monitoring the integrity of the estimate of the position of a mobile carrier in a satellite positioning measurement system
    EP2339377A1|2011-06-29|Satellite positioning receiver
    EP1813957B1|2010-03-10|Device controlling relative position| by analysing dual frequency signals for use by a spacecraft in a group of spacecraft in formation
    EP2902797B1|2021-07-21|Radio-frequency method and system for determining, using spacecraft torque, the relative angular position between a plurality of remote spacecraft
    WO2010034800A1|2010-04-01|Method for optimising an acquisition of a spread-spectrum signal from a satellite by a mobile receiver
    EP2500750B1|2017-11-22|Method and device for calibrating a receiver.
    EP2487507B1|2017-12-20|Acquisition method with improved frequency multi-correlation
    EP2221628B1|2012-05-09|Method for removing ambiguity, method for positioning a receiver by radionavigation including a step of removing ambiguity and positioning receiver
    EP2642317B1|2020-05-06|Device for receiving radio-navigation signals with multiple antennas
    WO2002033437A1|2002-04-25|System for tracking a moving object
    EP2998765B1|2017-03-29|System for excluding a failure of a satellite in a gnss system
    CA2805068C|2019-12-31|Determination process for a confidence indicator relative to a trajectory followed by a mobile
    EP2642319B1|2020-04-22|Device for receiving radio-navigation signals with multiple antennas and common synchronisation control
    EP1752786B1|2013-11-20|Hybrid inertial/satellite navigation system and method for the operation of such a system
    EP3353573B1|2020-03-18|Railway positioning method and system
    EP2476015B1|2018-10-31|Method for determining the navigation speed of a carrier and hybridization device
    Charbonnieras2017|Exploitation of the GNSS signals for integrity measurement.
    EP2687869B1|2017-03-15|Method for acquiring GNSS signals for a satellite and receiver suitable for implementing said method
    WO2021255386A2|2021-12-23|Receiver of radionavigation signals comprising a computer of a correlation power indicator
    WO2020025518A1|2020-02-06|Method and device for locating a moving object subjected to rotational dynamics on its own axis
    同族专利:
    公开号 | 公开日
    EP2921923A1|2015-09-23|
    US20150247730A1|2015-09-03|
    US10054449B2|2018-08-21|
    FR3018121B1|2016-02-12|
    EP2921923B1|2018-07-11|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US6859170B2|1999-04-21|2005-02-22|The Johns Hopkins University|Extended kalman filter for autonomous satellite navigation system|
    US6882908B2|2000-09-15|2005-04-19|Centre National D'etudes Spatiales|Method for calculating instantaneous characteristics of a satellite in orbit, equipped with a GNSS receiver|
    US20110254734A1|2010-04-14|2011-10-20|The Boeing Company|Software GNSS Receiver for High-Altitude Spacecraft Applications|
    US7370834B2|1993-11-12|2008-05-13|The Baron Company, Ltd.|Apparatus and methods for in-space satellite operations|
    US5595360A|1994-03-25|1997-01-21|Hughes Aircraft Company|Optimal transfer orbit trajectory using electric propulsion|
    US6845950B1|2003-11-26|2005-01-25|Lockheed Martin Corporation|System for high efficiency spacecraft orbit transfer|
    US7246775B1|2004-08-02|2007-07-24|Lockheed Martin Corporation|System and method of substantially autonomous geosynchronous time-optimal orbit transfer|
    WO2008147232A1|2007-05-29|2008-12-04|Deimos Engenharia S.A.|Highly integrated gps, galileo and inertial navigation system|FR3043985B1|2015-11-20|2018-04-20|Thales|METHOD FOR ORBITAL TRANSFER OF A SPATIAL VESSEL USING CONTINUOUS OR QUASI-CONTINUOUS THRUST AND ON-LINE STEERING SYSTEM FOR IMPLEMENTING SUCH A METHOD|
    CN106295218B|2016-08-19|2019-04-19|北京航空航天大学|A kind of quick determining energetic optimum intercepts the numerical optimization of predicted set-forward position|
    CN108344415B|2018-01-30|2021-06-04|北京大学|Combined navigation information fusion method|
    US10650687B2|2018-05-18|2020-05-12|Aireon Llc|Decoding position information in space-based systems|
    CN109005108A|2018-06-29|2018-12-14|全球能源互联网研究院有限公司|A kind of spatial network route planning method and system|
    US10979876B2|2018-08-31|2021-04-13|Cohda Wireless Pty Ltd.|Method for estimating the position of an object|
    CN109625332B|2018-12-29|2022-03-01|西北工业大学|Preset performance control method for track crossing of translation point without initial error symbol|
    CN109774974B|2019-01-31|2020-08-14|上海微小卫星工程中心|Track design method for space debris approaching|
    CN111942618B|2020-07-08|2022-01-04|北京控制工程研究所|GNSS data-based track acquisition method suitable for in-motion imaging|
    法律状态:
    2015-02-10| PLFP| Fee payment|Year of fee payment: 2 |
    2016-01-25| PLFP| Fee payment|Year of fee payment: 3 |
    2017-01-26| PLFP| Fee payment|Year of fee payment: 4 |
    2018-01-26| PLFP| Fee payment|Year of fee payment: 5 |
    2019-01-25| PLFP| Fee payment|Year of fee payment: 6 |
    2020-11-06| ST| Notification of lapse|Effective date: 20201006 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1400508A|FR3018121B1|2014-02-28|2014-02-28|METHOD FOR TRACKING A TRANSFER ORBIT OR A PHASE FOR ORKING A SPATIAL VEHICLE, IN PARTICULAR AN ELECTRICAL PROPULSION, AND APPARATUS FOR IMPLEMENTING SUCH A METHOD|FR1400508A| FR3018121B1|2014-02-28|2014-02-28|METHOD FOR TRACKING A TRANSFER ORBIT OR A PHASE FOR ORKING A SPATIAL VEHICLE, IN PARTICULAR AN ELECTRICAL PROPULSION, AND APPARATUS FOR IMPLEMENTING SUCH A METHOD|
    EP15156676.7A| EP2921923B1|2014-02-28|2015-02-26|Method for tracking a transfer orbit or a phase of placing a space vehicle in orbit, in particular a vehicle with electric drive, and apparatus for implementing such a method|
    US14/633,512| US10054449B2|2014-02-28|2015-02-27|Method of following a transfer orbit or a phase of orbital placement of a space vehicle, in particular an electric propulsion vehicle, and apparatus for the implementation of such a method|
    [返回顶部]