Starlight navigation convergence optimization method
By designing a multi-field composite starlight observation sensor and establishing a joint observation model, combining time-sharing navigation strategies and optimized dynamic models, the problem of slow convergence speed of deep space autonomous navigation is solved, and efficient and reliable deep space navigation is achieved.
Patent Information
- Application Number
- CN202510299005.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-13
- Publication Date
- 2025-06-13
AI Technical Summary
In the prior art, deep space autonomous navigation based on the optical difference effect converges slowly in a deep space environment, affecting the smooth progress of the task.
A multi-field composite starlight observation and angle distance observation are designed to integrate observation sensor configuration. By establishing a starlight navigation observation model based on multi-field angle distance measurement and a deep space navigation observation model based on large celestial line of sight vector and center distance, a joint observation equation is formed, and a time-sharing navigation observation strategy and optimization dynamic model are designed according to the task characteristics, a rapid solution to the spacecraft navigation state quantity is achieved.
It effectively improves the convergence speed of deep space autonomous navigation, realizes high-precision, high reliability, continuous and robust autonomous navigation, and supports the high-precision navigation requirements of deep space missions.
Smart Images
Figure CN120141452A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of spacecraft autonomous navigation, in particular to a starlight navigation convergence optimization method. Background Art
[0002] Human space activities are gradually moving towards the distant deep space. In order to ensure that deep space missions can successfully achieve the intended results, higher requirements are placed on the reliability of space missions. Navigation data is crucial to ensuring mission reliability. Existing deep space mission navigation mainly relies on ground measurement and control, which makes distant deep space missions have significant time delay characteristics, affecting the smooth progress of the mission. At present, aerospace-related institutions in various countries have successively carried out a large number of autonomous navigation system program research. The proposed solutions cover the use of various types of satellite-borne equipment such as sun sensors, earth sensors, star sensors, magnetometers, X-ray detectors, optical cameras, etc. for autonomous navigation, and have carried out the development of corresponding navigation systems and satellite-borne equipment. Among various autonomous navigation methods, the autonomous navigation method based on optical imaging has high adaptability and mature data acquisition technology. It is the most widely demonstrated and used deep space autonomous navigation method.
[0003] Existing deep space optical autonomous navigation methods mainly use imaging of the sun, large planets, and asteroids to achieve navigation, but the navigation sources are limited and the spatial distribution density is low and uneven, which imposes certain restrictions on spacecraft observations. Background stars in space are widely distributed in the celestial background. In the past, imaging of background stars was used for attitude determination. With the improvement of observation capabilities, it has become possible to use the relativistic aberration effect in star imaging for navigation, which solves the problems caused by limited and uneven distribution of navigation sources and can achieve effective navigation in the entire deep space. Autonomous navigation using the aberration effect is, in principle, to image stars to invert the spacecraft speed. This autonomous navigation method based on spacecraft velocity observations has the characteristic of slow convergence in deep space environments, which affects the deployment of full-stage missions. Therefore, how to optimize the convergence speed of aberration navigation in deep space environments is an urgent problem to be solved. Summary of the invention
[0004] The technical problem solved by the present invention is: to overcome the deficiencies of the prior art, to provide a starlight navigation convergence optimization method, and to solve the problem of slow convergence speed of deep space autonomous navigation using the aberration effect.
[0005] The technical solution of the present invention is: a starlight navigation convergence optimization method, comprising:
[0006] Establish an observation sensor configuration that integrates multi-field composite starlight observation and angle and distance measurement observation; this configuration includes multiple starlight navigation fields for star angular distance observation, and one field for large celestial body imaging observation to obtain the line of sight vector and center distance of large celestial bodies;
[0007] Through the above configuration, a starlight navigation observation model based on multi-field-of-view angular distance measurement and a deep space navigation observation model based on the line-of-sight vector and central distance of large celestial bodies are established to form a joint observation equation;
[0008] According to the mission characteristics, a time-sharing navigation observation strategy is designed for the joint observation equation;
[0009] According to the deep space working conditions, an optimized dynamic model considering the fast convergence characteristics is constructed;
[0010] According to the time-sharing navigation observation strategy, using the established joint observation equation and the state equation generated by the optimized dynamic model, the rapid solution of the spacecraft navigation state quantity is realized.
[0011] Furthermore, the observation sensor configuration is a four-field-of-view joint observation sensor configuration, where the optical axes of three starlight navigation fields of view are mutually at an angle of 90°, and the fourth field of view for imaging large celestial bodies is located in the center of the three fields of view.
[0012] Furthermore, the specific method for establishing the starlight navigation observation model based on multi-field-of-view angular distance measurement is as follows:
[0013] Stars within their respective fields of view are extracted using three starlight navigation fields of view, and at least three pairs of star angular distances are obtained by pairwise combination using the description of stellar aberration explained by special relativity as direct observables, and a mathematical correlation between the star angular distances and the spacecraft navigation state quantities is established.
[0014] Furthermore, at least one star with a signal-to-noise ratio better than magnitude 8 is identified and selected in each starlight navigation field of view.
[0015] Furthermore, the specific method for establishing the deep space navigation observation model based on the line-of-sight vector and central distance of large celestial bodies is as follows:
[0016] The center of the large celestial body is extracted from the imaging result of the large celestial body, the line-of-sight vector of the large celestial body is calculated, and the central distance of the spacecraft relative to the large celestial body is estimated according to the size of the imaging area of the large celestial body field of view and the radius of the large celestial body; with the line-of-sight vector and central distance of the large celestial body as observables, a mathematical correlation with the celestial ephemeris and the spacecraft navigation state quantities is established.
[0017] Furthermore, the specific method for designing the time-sharing navigation observation strategy for the joint observation equation according to the mission characteristics is as follows:
[0018] The joint observation equation is an equation containing 7-dimensional observables (θ 1 , θ 2 , θ 3 , e, ρ) and the 6-dimensional orbital state quantities of the spacecraft orbit (x, y, z, v x , v y , v z) Observation equation; the time-sharing navigation observation strategy is as follows: full-dimensional observation is adopted at the start of the mission and when approaching a large celestial body, and the observation quantities are (θ 1 , θ 2 , θ 3 , e, ρ). When far from the large celestial body, only the stellar aberration observation is used, and the observation quantities degenerate to (θ 1 , θ 2 , θ 3 ); where θ 1 , θ 2 , θ 3 are the three stellar angular distances obtained by pairwise combination of three starlight navigation fields of view, e, ρ are the line-of-sight vector and the central distance of the large celestial body, x, y, z are the three-dimensional position vectors of the spacecraft, and v x , v y , v z are the three-dimensional velocity vectors of the spacecraft.
[0019] Further, the optimization dynamics model considering the fast convergence characteristics is constructed according to the deep space working conditions, and the specific method is as follows:
[0020] In the cruise section, the heliocentric inertial reference frame is used for description, and the disturbing forces including the solar two-body gravity, the gravitational perturbation of large celestial bodies in the solar system, the light pressure, and the relativistic effect are considered to establish a dynamics model; in the celestial body orbiting section, in addition to the above disturbing forces, the non-spherical gravity of the central celestial body and the atmospheric drag of the celestial body are additionally considered to establish a dynamics model.
[0021] Further, the fast solution of the spacecraft navigation state quantity is as follows:
[0022] By establishing a sequential distributed processing framework, the distributed federated extended Kalman filter algorithm is used for solution; a total of two sub-filters are formed. One sub-filter is used to process the starlight navigation observation data, and the other sub-filter is used to process the large celestial body observation data. According to the criterion of whether it is close to the large celestial body, it is judged whether to turn on the sub-filter for large celestial body observation, and finally the real-time estimation of the spacecraft navigation state quantity is realized.
[0023] The present invention also provides a computer program product, and when the computer program product is executed by a processor, the steps of the method described above are implemented.
[0024] A computing terminal of the present invention is deployed on a deep space detector and includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method described above are implemented.
[0025] The advantages of the present invention compared with the prior art are as follows:
[0026] (1) The prior art has not analyzed the deep space autonomous navigation method from the perspective of navigation convergence, especially the convergence speed. Autonomous navigation based on the observation of the starlight angular distance of stellar aberration has the characteristics of high precision and high stability, but the navigation convergence speed is limited. The present invention proposes an optimization method for effectively improving the autonomous navigation convergence speed from the perspectives of sensor configuration and navigation solution method. By designing a multi-field-of-view observation sensor configuration, starlight navigation assisted by large celestial body imaging ranging and angle measurement is carried out. By introducing directly position-related observation data and more reasonably describing the dynamics, rapid convergence of the navigation position and speed is achieved.
[0027] (2) The present invention can strongly support the practical feasibility of deep space high-precision, high-reliability, continuous and robust autonomous navigation methods. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 is a flowchart of an optimization method for the starlight navigation convergence performance of the present invention;
[0029] Figure 2 is a configuration diagram of a multi-field-of-view optical imaging navigation sensor for starlight navigation and large celestial body observation. DETAILED DESCRIPTION OF THE INVENTION
[0030] In order to better understand the technical solution of the present invention, the specific implementation manner of the present invention will be described below.
[0031] As Figure 1 shown, an optimization method for starlight navigation convergence proposed by the present invention includes the following steps:
[0032] Step 1: Establish an observation sensor configuration that integrates multi-field-of-view composite starlight observation and angle measurement and ranging observation.
[0033] Design a multi-field-of-view starlight navigation sensor configuration. Each field of view independently images, obtains the background star map and matches it with the nominal star ephemeris table to obtain the attitude of each field of view relative to the inertial space. Taking the inertial space as a medium, the attitude conversion relationship between each field of view can be obtained. At least one star with a signal-to-noise ratio better than magnitude 8 is identified and selected from each field of view. Stars in different fields of view can form an inter-star angular distance observation with a large angular distance through the attitude matrix. In order to improve the navigation convergence characteristics, on this basis, a large celestial body imaging observation field of view is specially designed, which is located in the middle of the three fields of view of starlight navigation and can achieve sub-arcsecond imaging observation of large celestial bodies.
[0034] Specifically, as Figure 2As shown in the figure, in this embodiment, a four-field-of-view combined observation sensor configuration is established, where the optical axes of the three fields of view of the starlight navigation are mutually at an angle of 90°. It is used for the large star angular distance observation required for starlight navigation based on stellar observation. According to the starlight navigation observation characteristics based on the aberration effect, the inter-star angular distance with a large separation angle can effectively improve the navigation accuracy; the fourth field of view is located in the center of the three fields of view of the starlight navigation, and the central large field of view is used for imaging observation of large celestial bodies (including the main planets within the solar system) to obtain the line-of-sight vector and central distance of the large celestial bodies.
[0035] Step 2: Establish a starlight navigation observation model based on multi-field-of-view angular distance measurement and a deep space navigation observation model based on the line-of-sight vector and central distance of large celestial bodies to form a combined observation equation.
[0036] The combined observation equation is an observation equation that includes a 7-dimensional observation quantity (θ 1 , θ 2 , θ 3 , e, ρ) and the 6-dimensional orbital state quantity (x, y, z, v x , v y , v z ) of the spacecraft orbit. The following gives a preferred method for establishing the combined observation equation:
[0037] (1) Establish a starlight navigation observation model based on multi-field-of-view angular distance measurement
[0038] First, perform star chart matching in the three starlight navigation fields of view. According to the relative geometric relationship of the stars within the fields of view, use the measured optical images of the celestial background stars obtained in each field of view and the pre-constructed star catalog, and perform matching based on the triangle matching algorithm to obtain the nominal line-of-sight directions of the stars (u 1 , u 2 ,... u i ...) in the image.
[0039] In a possible implementation, perform image preprocessing on the measured optical images obtained in orbit to improve the signal-to-noise ratio. Pre-construct a star catalog based on the GAIA star catalog according to the capabilities of the sensor and the requirements for the nominal accuracy of the observed stars. Use the preprocessed measured optical images and the star catalog, and complete the matching using a matching algorithm that constructs the triangles of the stars within the field of view as geometric features to obtain the nominal line-of-sight directions of the stars (u 1 , u 2 ,... u i ...) in the measured optical image.
[0040] Then, according to the description of stellar aberration explained by the special theory of relativity, based on the nominal line-of-sight directions of the stars, obtain the theoretical line-of-sight directions of the stars after the influence of aberration. For the theoretical line-of-sight directions of the stars after the influence of aberration of the three stars, perform inner product pairwise to obtain three star angular distances (θ 1, θ 2 , θ 3 , …); In order to eliminate the errors introduced by the absolute attitude of the starlight sensor, the star angular distance is used to replace the star line-of-sight measurement, a mathematical correlation between the observed angular distance and the spacecraft navigation state variables is established, and a starlight navigation observation model based on multi-field-of-view angular distance measurement is obtained.
[0041] In a possible implementation, the stars observed in three fields of view are resolved respectively. According to the Lorentz principle of special relativity, the theoretical star line-of-sight direction affected by aberration is constructed:
[0042]
[0043] where u ′ is the theoretical star line-of-sight direction affected by aberration, u is the nominal star line-of-sight direction, and γ is the Lorentz factor.
[0044] (2) Establish a deep space navigation observation model based on the line-of-sight vector and central distance of large celestial bodies
[0045] Point the middle field of view at a nearby large celestial body. At the initial moment of the deep space mission, point it in the direction of the Earth or the Moon. Extract the center of the large celestial body from the optical image, calculate the direction of the center in the spacecraft body direction, and estimate the distance ρ between the spacecraft and the large celestial body through the radius R of the large celestial body and the imaging radius of the focal plane.
[0046]
[0047] where α is the imaging field-of-view angle corresponding to the diameter of the large celestial body.
[0048] The finally constructed joint observation equation includes 7-dimensional observables, and the observables are (θ 1 , θ 2 , θ 3 , e, ρ). The joint observation model is as follows:
[0049]
[0050] where the subscripts i, j, k represent different stars, e, ρ are the line-of-sight vector and central distance of the large celestial body, r is the position vector of the spacecraft (i.e., x, y, z), r b is the position vector of the large celestial body, is the ratio of the spacecraft velocity vector v (i.e., v x , v y , v z ) to the speed of light c.
[0051] Step 3: According to the mission characteristics, design a time-sharing navigation observation strategy for the joint observation equation.
[0052] According to the mission process, full-dimensional observations are performed when the mission starts and when approaching a large celestial body to accelerate navigation convergence. When far from the large celestial body, only the stellar aberration observation is used, and the observation quantity degenerates to (θ 1 , θ 2 , θ 3 ).
[0053] Step 4: According to the deep-space working conditions, construct an optimized dynamic model considering the characteristics of rapid convergence.
[0054] To a certain extent, the dynamic model in navigation solution affects the navigation convergence performance. To accelerate the navigation convergence speed, the heliocentric inertial reference system is used to describe in the cruise section. The dynamics considers perturbations such as the solar two-body gravity, the gravitational perturbation of large celestial bodies in the solar system, the light pressure, and the relativistic effect. In the celestial body orbiting section, in addition to the above-mentioned perturbation forces, the non-spherical gravity of the central celestial body and the atmospheric drag of the celestial body are additionally considered. By considering a relatively complete dynamic model, the constraint of the dynamic model on the solution of the orbital state quantity can be increased, and the convergence performance can be improved.
[0055] Step 5: According to the time-sharing navigation observation strategy, use the established joint observation equation and the state equation generated by the optimized dynamic model to realize the rapid solution of the navigation state quantity of the spacecraft.
[0056] To disperse and reduce the computational burden, and for fault diagnosis and isolation, and at the same time adapt to the current situation of the two types of observations in this scheme, based on the joint observation equation in Step 2 and the state equation generated by the optimized dynamic model in Step 4, by establishing a sequential distributed processing framework, the distributed federated extended Kalman filter algorithm is used for processing, and finally the real-time estimation of the navigation state quantity is realized. In the distributed processing framework, two sub-filters are formed. One sub-filter is used to process the starlight navigation observation data, and the other sub-filter is used to process the large celestial body observation data. According to the criterion of whether it is close to the large celestial body, it is judged whether to turn on the sub-filter for large celestial body observation, and finally the real-time estimation of the navigation state quantity of the spacecraft is realized.
[0057] It can be understood that the present invention is described by way of examples. Those skilled in the art know that without departing from the spirit and scope of the present invention, various changes or equivalent substitutions can be made to these features and examples. In addition, under the teaching of the present invention, these features and examples can be modified to adapt to specific situations without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and the embodiments that can fall within the scope of the claims of this application all belong to the scope protected by the present invention.
[0058] The content not detailedly described in the specification of the present invention belongs to the well-known technology of those skilled in the art.
Claims
1. A starlight navigation convergence optimization method, characterized in that: include: Establish an observation sensor configuration that integrates multi-field composite starlight observation and angle and distance measurement observation; this configuration includes multiple starlight navigation fields for star angular distance observation, and one field for large celestial body imaging observation to obtain the line of sight vector and center distance of large celestial bodies; Through the above configuration, a starlight navigation observation model based on multi-field-of-view angular distance measurement and a deep-space navigation observation model based on large celestial body sight vector and center distance are established to form a joint observation equation; According to the mission characteristics and the joint observation equation, a time-sharing navigation observation strategy is designed; According to the deep space working conditions, an optimization dynamics model considering fast convergence characteristics is constructed; According to the time-sharing navigation observation strategy, the established joint observation equation and the state equation generated by the optimized dynamics model are used to quickly solve the spacecraft navigation state quantity.
2. The starlight navigation convergence optimization method according to claim 1, characterized in that: The observation sensor configuration is a four-field-of-view combined observation sensor configuration, wherein the optical axes of the three starlight navigation fields are at an angle of 90° to each other, and the fourth field of view used for large celestial body imaging observation is located in the center of the three fields of view.
3. The starlight navigation convergence optimization method according to claim 2, characterized in that: The starlight navigation observation model based on multi-field-of-view angular distance measurement is established in the following specific manner: The three starlight navigation fields are used to extract the stars in their respective fields of view. The stellar aberration described by the special theory of relativity is used to obtain at least three pairs of stellar angular distances as direct observation quantities by combining them two by two, and the mathematical relationship between the stellar angular distances and the spacecraft navigation state quantities is established.
4. The starlight navigation convergence optimization method according to claim 3, characterized in that: For each starlight navigation field, at least one star with a high signal-to-noise ratio better than magnitude 8 shall be identified and selected.
5. The starlight navigation convergence optimization method according to claim 1, characterized in that: The deep space navigation observation model based on the line of sight vector and center distance of a large celestial body is established in the following specific manner: The center of the large celestial body is extracted from the imaging results of the large celestial body, and the line of sight vector of the large celestial body is calculated. According to the size of the imaging area of the large celestial body's field of view and the radius of the large celestial body, the center distance of the spacecraft relative to the large celestial body is estimated; with the line of sight vector and the center distance of the large celestial body as the observed quantities, a mathematical relationship with the celestial body ephemeris and the spacecraft navigation state quantity is established.
6. The starlight navigation convergence optimization method according to claim 1, characterized in that: According to the characteristics of the task, a time-sharing navigation observation strategy is designed for the joint observation equation. The specific method is as follows: The joint observation equation contains the 7-dimensional observation quantities (θ1, θ2, θ3, e, ρ) and the 6-dimensional orbital state quantities (x, y, z, v x ,v y ,v z ) observation equation; the time-sharing navigation observation strategy is: full-dimensional observation is used at the beginning of the mission and when approaching a large celestial body, and the observed quantity is (θ1, θ2, θ3, e, ρ); when far away from a large celestial body, only stellar aberration observation is used, and the observed quantity degenerates into (θ1, θ2, θ3); where θ1, θ2, θ3 are the three stellar angular distances obtained by combining the three starlight navigation fields in pairs, e, ρ are the line of sight vector and the center distance of the large celestial body, x, y, z are the three-dimensional position vectors of the spacecraft, and v x ,v y ,v z is the three-dimensional velocity vector of the spacecraft.
7. The starlight navigation convergence optimization method according to claim 1, characterized in that: According to the deep space working conditions, an optimization dynamics model considering fast convergence characteristics is constructed, and the specific method is as follows: During the cruise phase, the heliocentric inertial reference system is used to describe the mission, and the perturbations including the gravity of the two suns, the gravitational perturbations of large celestial bodies in the solar system, light pressure, and relativistic effects are taken into account to establish a dynamic model. During the celestial body orbit phase, in addition to the above perturbations, the non-spherical gravity of the central celestial body and the atmospheric drag of the celestial body are additionally considered to establish a dynamic model.
8. The starlight navigation convergence optimization method according to claim 1, characterized in that: The specific method for quickly solving the spacecraft navigation state quantity is as follows: By establishing a sequential distributed processing framework, a distributed federated extended Kalman filter algorithm is used for solving the problem. Two sub-filters are formed, one sub-filter is used to process starlight navigation observation data, and the other sub-filter is used to process large celestial body observation data. The sub-filter for large body observation is determined based on whether the spacecraft is close to a large celestial body, thereby realizing real-time estimation of the navigation state of the spacecraft.
9. A computer program product, characterized in that: When the computer program product is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.
10. A computing terminal deployed in a deep space probe, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method as claimed in any one of claims 1 to 8 when executing the computer program.