Optimal selection method for optical navigation information of deep space probe

By constructing an optimized selection method for optical navigation information that integrates system observability and observation accuracy, and by using Fisher information matrix and genetic algorithm to optimize angle measurement combination, the problem of poor observability of angle measurement information of deep space probes was solved, and high-precision navigation and positioning was achieved.

CN118913246APending Publication Date: 2024-11-08BEIJING INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410950829.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Existing navigation and positioning methods for deep space probes suffer from poor observability of angle information selection results in angle measurement and ranging combined positioning, resulting in low navigation and positioning accuracy. Furthermore, existing methods introduce truncation errors or high-order nonlinear systems during linearization, leading to large computational loads and making it difficult to achieve optimized selection of online navigation information.

Method used

A method for optimizing the selection of optical navigation information based on the overall system observability and observation accuracy is constructed. The Fisher information matrix is ​​used to evaluate the overall observability of angle measurement information. A combined positioning scheme of optical angle measurement and navigation satellite ranging is designed. The optimal angle measurement combination is selected through a genetic algorithm, and a linearized combined positioning equation is derived.

Benefits of technology

It has achieved autonomous and precise positioning without navigation satellite ranging information, and high-precision combined positioning with ranging information, thus improving the navigation and positioning accuracy of deep space probes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913246A_ABST
    Figure CN118913246A_ABST
Patent Text Reader

Abstract

The invention discloses an optimal selection method for optical navigation information of a deep space probe, and belongs to the technical field of deep space exploration. The method comprises the following steps: aiming at the problem of optimal selection of angle measurement and distance measurement combined positioning optical observation information of a deep space probe, constructing an optical navigation information optimal selection method integrating system observability and observation precision, designing an optical angle measurement-navigational star distance measurement combined positioning scheme, and deducing a linearized combined positioning equation. The overall observability of angle measurement information is evaluated by utilizing a Fisher information matrix, the relative observability of the angle measurement information is designed by considering the observation precision of different sensors, and the overall observability and the relative observability are weighted and summed to construct an optical angle measurement information evaluation function of combined positioning. And taking the evaluation function as an optimization index, and selecting an optimal angle measurement combination by using a genetic algorithm. And the distance measurement information and the optimized angle measurement information are used as observed quantities, and high-precision positioning of the deep space probe is realized by using a batch processing algorithm. The method is particularly suitable for positioning the deep space probe on a transfer track or a long-distance approaching track.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for optimizing and selecting optical navigation information of a deep space detector, which is particularly suitable for positioning the deep space detector in a transfer orbit or a long-distance approach orbit and belongs to the technical field of deep space detection. Background Art

[0002] Deep space probes need to be positioned during flight to keep the probes on the predetermined orbit, provide support for the development of scientific missions, and adjust the antenna pointing according to the positioning results to maintain the probe's communication with the ground. Due to the long flight time, long distance from the earth, and complex space environment, deep space probes can use highly autonomous and real-time optical navigation to achieve their own positioning. In order to reduce the positioning error caused by factors such as ephemeris error and near-celestial body measurement accuracy, the ranging information of the navigation star can be introduced for combined navigation. At the same time, the number of navigation objects captured by the optical camera will change over time, and different observation configurations will affect the performance of the navigation system. In existing missions, probes usually track specific celestial bodies for navigation and positioning, but the selected celestial bodies do not necessarily have the best observability, which in turn affects the navigation performance. Therefore, evaluating and selecting the optical observation information in combined navigation can further improve the positioning accuracy.

[0003] Among the methods for optimizing navigation information for deep space probes that have been developed, the prior art [1] (see Lévesque JF, Lafontaine J D. Innovative Navigation Schemes for State and Parameter Estimation During Mars Entry [J]. Journal of Guidance, Control, and Dynamics, 2007, 30 (1): 169-184.) directly linearizes the nonlinear system to construct the observability matrix of the navigation system, and combines the theory of matrix canonical decomposition to analyze the impact of the amount of navigation information on the observability of the navigation system in the Mars atmosphere entry scenario, and gives a navigation information optimization scheme. This method will introduce truncation errors in the linearization process, and cannot guarantee that the observability of the linearized system is globally consistent with the original nonlinear system.

[0004] Prior art [2] (see Kassas ZM, Humphreys TE. Observability analysis of opportunistic navigation with pseudorange measurements. Proceedings of AIAA Guidance, Navigation, and Control Conference [C]. Red Hook, NY: Curran Associates, Inc, 2012, 2959-2974.) uses high-order Lie derivatives to approximate linear time-varying systems, introduces piecewise steady-state system theory to calculate the overall observability matrix of the navigation system, and uses this to simulate and analyze the performance of ground radio navigation with some known states. This method is mostly used for theoretical analysis because the calculation of Lie derivatives of high-order nonlinear systems is large, and it is difficult to achieve online optimization of navigation information selection. Summary of the invention

[0005] In order to solve the problem of poor observability of angle measurement information selection results in angle measurement and ranging combined positioning, resulting in low navigation positioning accuracy, the purpose of the present invention is to provide a method for optimizing the selection of optical navigation information of deep space detectors. The method aims at optimizing the selection of optical observation information for angle measurement and ranging combined positioning of deep space detectors, constructs an optical navigation information optimization selection method that integrates system observability and observation accuracy, designs a combined positioning scheme of optical angle measurement-navigation star ranging, and derives a linearized combined positioning equation. The overall observability of angle measurement information is evaluated using the Fisher information matrix, the relative observability of angle measurement information is designed considering the observation accuracy of different sensors, and the weighted sum of the two is used to construct an optical angle measurement information evaluation function for combined positioning. The evaluation function is used as an optimization indicator, and a genetic algorithm is used to select the optimal angle measurement combination. The ranging information and the preferred angle measurement information are used as observation quantities, and an optimization method is used to achieve high-precision positioning of deep space detectors.

[0006] The purpose of the present invention is achieved through the following technical solutions.

[0007] The method for optimizing and selecting optical navigation information of a deep space probe disclosed in the present invention comprises the following steps:

[0008] Step 1: Use sensors to obtain reference celestial body information and construct angle measurement observation equations;

[0009] Step 2: Determine whether the detector can receive available ranging information, if yes, proceed to step 3, if not, proceed to step 4;

[0010] Step 3: Construct the ranging observation equation based on the available ranging information obtained in step 2;

[0011] Step 4: According to the angle measurement observation equation of step 1 and the distance measurement observation equation of step 3, construct an evaluation function for the angle measurement information; optimize the angle measurement information according to the evaluation function to obtain the optimal angle measurement combination, that is, to achieve the optimal selection of optical navigation information for deep space probes.

[0012] The specific implementation method of step 1 is:

[0013] The pixel coordinates p of the i-th reference celestial body from the camera's perspective i , Image line l i The coordinates are expressed as follows:

[0014]

[0015] Among them, c mn ,m=1,2,3n=1,2,3 is the coordinate transformation matrix C from the J2000.0 heliocentric ecliptic coordinate system to the detector body coordinate system cm The corresponding elements in the graph are x, y, and z, which are the position coordinates of the detector in the J2000.0 heliocentric ecliptic coordinate system. i ,y i , z i is the position coordinate of the reference celestial body in the J2000.0 heliocentric ecliptic coordinate system, and f is the focal length of the camera;

[0016] The observation angle obtained by the camera observing any two reference celestial bodies i and j is A ij , according to the geometric relationship, we have

[0017]

[0018] Among them, r i With r j is the reference celestial coordinate of the detector system, r i With r j r i With r j The scalar size of

[0019] The pixel and image line coordinates of the optical information obtained by equations (1) and (2) represent the observation angle, and we have:

[0020]

[0021] Using the vectors in the J2000.0 heliocentric ecliptic coordinate system to geometrically describe the observation angle, we have

[0022] (ρ-ρ i )(ρ-ρ j )=|ρ-ρ i ||ρ-ρ j |cos A ij(5) Where ρ is the detector position vector, ρ i , ρ j is the reference celestial position vector;

[0023] Linearizing equation (5) yields:

[0024]

[0025] Among them, δA is the angle deviation, δr is the position deviation;

[0026] Let the unit vector of the line of sight from the detector to the target object be Then formula (6) is transformed into

[0027]

[0028] Define auxiliary vector n ij With n ji

[0029]

[0030] Substituting formula (8) into formula (7), we have

[0031]

[0032] To further simplify the angle deviation δA ij The relationship between the position deviation δr is:

[0033]

[0034] The linearized observation equation is

[0035] δA ij =h ij ·δr (11)

[0036] The observation angle information of any two reference celestial bodies i and j is obtained by equation (4). According to the three pairs of information δA i1j1 , δA i2j2 and δA i3j3 (The observation angles used are: celestial body i 1 With celestial body 1 The angle between the celestial bodies 2 With celestial body 2 The angle between the celestial bodies 3 With celestial body 3 The angle between them can be solved by equation (11) to get r.

[0037] The specific implementation method of step 3 is:

[0038] At a certain moment, the detector G measured the iThe pseudorange information of , the following pseudorange observation equation is obtained:

[0039]

[0040] Among them, d k The navigation star S k Pseudorange information relative to detector G, v k The noise error generated by the observation includes the error of the detector signal receiving payload and the noise generated by the signal transmission during the process;

[0041] The nominal position coordinates of the detector relative to the J2000.0 heliocentric ecliptic coordinate system are (x 0 ,y 0 ,z 0 ), Taylor expansion of equation (12) near the nominal position of the detector is obtained:

[0042]

[0043] Among them, R k For the probe and the navigation star S k The real distance

[0044] Define auxiliary amount

[0045]

[0046] Let h s =[l k m k n k ],δd k =d k -R k , the linearized ranging observation equation is:

[0047] δd k =h s ·δr (15)

[0048] The specific implementation method of step 4 is:

[0049] Step 4.1, when there is no ranging information, directly optimize the obtained observation angle information;

[0050] Using Fisher information matrix to measure the observability of optical angle measurement information

[0051]

[0052] in, represents the state transfer matrix, which is h in formula (11) ij ; σ i is the noise standard deviation.

[0053] The measurement error variance and the trace of the Fisher information matrix satisfy

[0054]

[0055] Among them, λ 1 ,λ 2 ,λ 3 is the eigenvalue of the Fisher information matrix;

[0056] The eigenvalue of the Fisher information matrix is ​​used as an evaluation index to measure the overall observability of a set of angle measurement information.

[0057] σ 2 =tr(F) (18)

[0058] When there is no distance information, the evaluation function of angle information is σ = σ 2 ; By obtaining a set of optical angle measurement information with the largest σ value, the optimal selection of the observation angle can be achieved;

[0059] Step 4.2: When there is ranging information, since the positioning accuracy of ranging is much higher than that of optical angle measurement, the observation angle can be optimized based on the high-precision ranging information, and then the combined positioning method can be used to realize the positioning of the detector;

[0060] The relative observability of the ranging information in the three axes is expressed by the normalized relative coordinate difference between the navigation star and the detector:

[0061]

[0062]

[0063] Among them, OBS orbit is the relative coordinate difference between the navigation star and the detector, NOBS orbit is the normalized OBS orbit ;

[0064] Calculate the inverse matrix F of the Fisher information matrix -1 The square root of the diagonal elements of , we can get the lower bound of the estimated error of each axis; using f 11 ,f 22 ,f 33 Indicates F -1 The square root of the diagonal elements of f, normalized 11 ,f 22 ,f 33 The reciprocal of indicates the relative observability of the three axes.

[0065]

[0066] Among them, OBSrover NOBS is the vector composed of the inverse square roots of the diagonal elements of the inverse matrix of the optical information Fisher information matrix. rover is the normalized OBS rover ;

[0067] The relative observability of equation (18) and equation (22) is weighted and added to obtain a set of compensation conditions of the three axes of the angle measurement information for the distance measurement information:

[0068] NOBS=NOBS orbit +κNOBS rover (twenty three)

[0069] Among them, κ is the relative observability weighting coefficient;

[0070] Establish the evaluation index σ of the relative observability of angle measurement information 1

[0071]

[0072] Where min(NOBS) and max(NOBS) represent the minimum and maximum values ​​of the vector NOBS respectively; σ 1 Satisfy 0≤σ 1 ≤1, the larger the value, the more likely the selected angle measurement combination is to ensure that all three axes have a higher degree of observability;

[0073] Combining equation (24) and equation (18), the evaluation function of angle measurement information when distance measurement information is known is established:

[0074] σ=ασ 1 +βσ 2 (25)

[0075] Among them, α is the relative observability weighting coefficient, and β is the overall observability weighting coefficient. By obtaining a set of optical angle measurement information with the largest σ value, the optimization of the observation angle is achieved.

[0076] The method further includes step 5: using the optimal angle measurement combination solved in step 4, obtaining and solving the positioning equation for detector navigation.

[0077] Step 5.1: If there is no available navigation star, use the linearized angle measurement observation equation of formula (11), select three angle measurement information, and solve the equation group to achieve angle measurement positioning;

[0078] make Among them, Z is the angle measurement observation vector, and H is the corresponding angle measurement observation matrix.

[0079] The angle measurement observation equations are written as

[0080] Z=H·δr (26)

[0081] Among them, δr is the position offset, which includes the three-axis components △x, △y and △z, and δr=[△x,△y,△z].

[0082] Step 5.2, when the detector receives the angle measurement and distance measurement signals, the linearized observation equation is obtained by using equations (11) and (15), three angle measurement information and one distance measurement information are selected, and the combined navigation positioning is realized by solving the equation group;

[0083] make Among them, Z' is the angle measurement and ranging observation vector, and H' is the corresponding angle measurement and ranging observation matrix.

[0084] The combined observation equations are written as

[0085] Z'=H'·δr (27)

[0086] The prior estimate of the detector position r can be obtained by using the detector's orbit prediction or trajectory recursion + , use formula (4) to determine the predicted value A of the angle measurement ij + ,use Determine the predicted value R of the range i + ,(x k ,y k ,z k ) is the position coordinate of the navigation star relative to the J2000.0 heliocentric ecliptic coordinate system at this time, (x + ,y + ,z + ) is r + , calculate the observation matrix H or H', and obtain the observation matrix Z or Z' by subtracting the predicted value from the observed value;

[0087] The current position of the detector can be obtained by solving equation (26) or equation (27) using the least squares method.

[0088] r=r + +(H T H) -1 H T Z (28)

[0089] Beneficial effects:

[0090] 1. The method for optimizing the selection of optical navigation information of deep space probes disclosed in the present invention uses the rank of the Fisher information matrix to characterize the overall observability of the angular measurement information. When there is no navigation star ranging information, the relative observability of the angular measurement information is designed taking into account the observation accuracy of different sensors. The weighted sum of the two is used to construct an optical angular measurement information evaluation function for combined positioning, that is, the overall observability is used as the evaluation function, and a genetic algorithm is used to select the optimal angular measurement combination, so as to realize autonomous and precise positioning of the deep space probe.

[0091] 2. The present invention discloses a method for optimizing the selection of optical navigation information for deep space probes, constructs an optical navigation information optimization selection method that integrates system observability and observation accuracy. When there is ranging information, the relative observability description equation of the angular measurement information is established based on the ranging information, an evaluation function is constructed in combination with the overall observability, and a genetic algorithm is used to solve the problem of maximizing the evaluation function value, thereby realizing the selection of the optimal angular measurement combination under known ranging information. The ranging information and the preferred angular measurement information are used as observation quantities, a combined positioning scheme of optical angular measurement-navigation star ranging is designed, a linearized combined positioning equation is derived, and a least squares algorithm is used to realize high-precision positioning of the deep space probe. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 It is a flow chart of the method for optimizing the selection of optical navigation information of a deep space probe disclosed in the present invention;

[0093] Figure 2 is the result of the Monte Carlo simulation of the three-axis position error under the angle measurement positioning of the example of the present invention, wherein (a) is the XY axis position error ellipse, (b) is the YZ axis position error ellipse, and (c) is the ZX axis position error ellipse;

[0094] Figure 3 It is the result of the Monte Carlo simulation of the three-axis position error under the combined positioning of the example of the present invention, where (a) is the XY axis position error ellipse, (b) is the YZ axis position error ellipse, and (c) is the ZX axis position error ellipse. DETAILED DESCRIPTION

[0095] In order to better illustrate the purpose and advantages of the present invention, the invention is further described below with reference to the accompanying drawings and examples.

[0096] In order to verify the feasibility of the present invention, a simulation is performed by taking Mars approaching orbit as an example. The probe is located at [-3349050.00, -10307316.05, 7874088.80] m, and the initial position deviation of the three axes of the probe is 5000 m.

[0097] like Figure 1 As shown, the method for optimizing the selection of optical navigation information of a deep space probe disclosed in this embodiment has the following specific implementation steps:

[0098] Step 1: Use sensors to obtain reference celestial body information and construct angle measurement observation equations;

[0099] The number of reference objects selected is 5 nearby objects and 15 distant objects.

[0100] The pixel coordinates p of the i-th reference celestial body from the camera's perspective i , Image line l i The coordinates are expressed as follows:

[0101]

[0102] Among them, c mn ,m=1,2,3n=1,2,3 is the coordinate transformation matrix C from the J2000.0 heliocentric ecliptic coordinate system to the detector body coordinate system cm The corresponding elements in the graph are x, y, and z, which are the position coordinates of the detector in the J2000.0 heliocentric ecliptic coordinate system. i ,y i , z i is the position coordinate of the reference celestial body in the J2000.0 heliocentric ecliptic coordinate system, and f is the focal length of the camera;

[0103] The observation angle obtained by the camera observing any two reference celestial bodies i and j is A ij , according to the geometric relationship, we have

[0104]

[0105] Among them, r i With r j is the reference celestial coordinate of the detector system, r i With r j r i With r j The scalar size of

[0106] The pixel and image line coordinates of the optical information obtained by equations (1) and (2) represent the observation angle, and we have:

[0107]

[0108] Using the vectors in the J2000.0 heliocentric ecliptic coordinate system to geometrically describe the observation angle, we have

[0109] (ρ-ρ i )(ρ-ρ j )=|ρ-ρ i ||ρ-ρ j |cos A ij (5) Where ρ is the detector position vector, ρ i , ρj is the reference celestial position vector;

[0110] Linearizing equation (5) yields:

[0111]

[0112] Among them, δA is the angle deviation, δr is the position deviation;

[0113] Let the unit vector of the line of sight from the detector to the target object be Then formula (6) can be transformed into

[0114]

[0115] Define auxiliary vector n ij With n ji

[0116]

[0117] Substituting formula (8) into formula (7), we have

[0118]

[0119] To further simplify the angle deviation δA ij The relationship between the position deviation δr is:

[0120]

[0121] The linearized observation equation is

[0122] δA ij =h ij ·δr (11)

[0123] Among them, δr is the position offset, which includes the three-axis components △x, △y and △z, and δr=[△x,△y,△z].

[0124] The observation angle information of any two reference celestial bodies i and j is obtained by equation (4). and (The observation angles used are: celestial body i 1 With celestial body 1 The angle between the celestial bodies 2 With celestial body 2 The angle between the celestial bodies 3 With celestial body 3 The angle between them can be solved by equation (11) to obtain r;

[0125] Step 2: Determine whether the detector can receive available ranging information. If yes, proceed to step 3; if no, proceed to step 4. The navigation star is located at [-1158839.20, -3236512.38, 1610345.85] m;

[0126] Step 3: Construct the ranging observation equation based on the available ranging information obtained in step 2;

[0127] At a certain moment, the detector G measured the i The pseudorange information of the navigation star at this time is (x k ,y k ,z k ), the following pseudo-range observation equation is obtained:

[0128]

[0129] Among them, d k The navigation star S k Pseudorange information relative to detector G, v k The noise error generated by the observation includes the error of the detector signal receiving payload and the noise generated by the signal transmission during the process;

[0130] The nominal position coordinates of the detector relative to the J2000.0 heliocentric ecliptic coordinate system are (x 0 ,y 0 ,z 0 ), Taylor expansion of equation (12) near the nominal position of the detector is obtained:

[0131]

[0132] Among them, R k For the probe and the navigation star S k The real distance

[0133] Define auxiliary amount

[0134]

[0135] Let h s =[l k m k n k ], δd k =d k -R k , the linearized ranging observation equation is:

[0136] δd k =h s ·δr (15)

[0137] Considering the ranging information of the navigation star, the ranging error is 5m.

[0138] Step 4: According to the angle measurement observation equation of step 1 and the distance measurement observation equation of step 3, construct an evaluation function of the angle measurement information; optimize the angle measurement information according to the evaluation function.

[0139] Step 4.1, when there is no ranging information, directly optimize the obtained observation angle information;

[0140] Using Fisher information matrix to measure the observability of optical angle measurement information

[0141]

[0142] in, represents the state transfer matrix, which is h in formula (11) ij ; σ i is the noise standard deviation.

[0143] The measurement error variance and the trace of the Fisher information matrix satisfy

[0144]

[0145] Among them, λ 1 ,λ 2 ,λ 3 is the eigenvalue of the Fisher information matrix;

[0146] The eigenvalue of the Fisher information matrix is ​​used as an evaluation index to measure the overall observability of a set of angle measurement information.

[0147] σ 2 =tr(F) (18)

[0148] When there is no distance information, the evaluation function of angle information is σ = σ 2 ; By obtaining a set of optical angle measurement information with the largest σ value, the optimal selection of the observation angle can be achieved;

[0149] Considering that the amount of angle measurement information is generally very large, and three angle measurements need to be selected each time for evaluation function calculation, in order to balance computational efficiency and the global optimality of the solution, a genetic algorithm is used to solve the optimization problem.

[0150] The genetic algorithm parameters are set as follows: the initial population size G s =50, genetic generation number Gen = 50, crossover probability P c =0.8, mutation probability P a =0.1, the selection operation uses roulette and performs elite selection.

[0151] Step 4.2: When there is ranging information, since the positioning accuracy of ranging is much higher than that of optical angle measurement, the observation angle can be optimized based on the high-precision ranging information, and then the combined positioning method can be used to realize the positioning of the detector;

[0152] The relative observability of the ranging information in the three axes is expressed by the normalized relative coordinate difference between the navigation star and the detector:

[0153]

[0154] Among them, OBS orbit is the relative coordinate difference between the navigation star and the detector, NOBS orbit is the normalized OBS orbit ;

[0155] Calculate the inverse matrix F of the Fisher information matrix -1 The square root of the diagonal elements of , we can get the lower bound of the estimated error of each axis; using f 11 ,f 22 ,f 33 Indicates F -1 The square root of the diagonal elements of f, normalized 11 ,f 22 ,f 33 The reciprocal of indicates the relative observability of the three axes.

[0156]

[0157] Among them, OBS rover NOBS is the vector composed of the inverse square roots of the diagonal elements of the inverse matrix of the optical information Fisher information matrix. rover is the normalized OBS rover ;

[0158] The relative observability of equation (20) and equation (22) is weighted and added to obtain a set of compensation conditions of the three axes of the angle measurement information for the distance measurement information:

[0159] NOBS=NOBS orbit +κNOBS rover (twenty three)

[0160] Among them, κ is the relative observability weighting coefficient;

[0161] Establish the evaluation index σ of the relative observability of angle measurement information 1

[0162]

[0163] Where min(NOBS) and max(NOBS) represent the minimum and maximum values ​​of the vector NOBS respectively; σ 1Satisfy 0≤σ 1 ≤1, the larger the value, the more likely the selected angle measurement combination is to ensure that all three axes have a higher degree of observability;

[0164] Combining equation (18) and equation (24), the evaluation function of angle measurement information when distance measurement information is known is established:

[0165] σ=ασ 1 +βσ 2 (25)

[0166] Among them, α is the relative observability weighting coefficient, and β is the overall observability weighting coefficient. By obtaining a set of optical angle measurement information with the largest σ value, the optimal selection of the observation angle can be achieved;

[0167] In the combined positioning simulation with ranging information, the evaluation function coefficients α=1, β=10 are selected 13 .

[0168] Genetic algorithm is also used to solve the problem of optimizing the selection of angle measurement information in combined positioning with distance measurement information.

[0169] The genetic algorithm parameters are set as follows: the initial population size G s =50, genetic generation number Gen = 50, crossover probability P c =0.8, mutation probability P a =0.1, the selection operation uses roulette and performs elite selection.

[0170] Step 5: Using the optimal angle measurement combination solved in step 4, obtain and solve the positioning equation for detector navigation;

[0171] Step 5.1: If there is no available navigation star, use the linearized angle measurement observation equation of formula (11), select three angle measurement information, and solve the equation group to achieve angle measurement positioning;

[0172] make Among them, Z is the angle measurement observation vector, and H is the corresponding angle measurement observation matrix.

[0173] The angle measurement observation equations are written as

[0174] Z=H·δr (26)

[0175] The prior estimate of the detector position r can be obtained by using the detector's orbit prediction or trajectory recursion + , use formula (4) to determine the predicted value A of the angle measurement ij + , calculate the observation matrix H, and obtain the observation matrix Z by subtracting the predicted value from the observed value;

[0176] The current position of the detector can be obtained by solving equation (26) using the least squares method:

[0177] r=r + +(H T H) -1 H T Z (27)

[0178] In order to ensure that the simulation results are not interfered by the noise random numbers, the simulation without ranging information was performed 100 times of Monte Carlo shooting.

[0179] The average positioning errors of the three axes of angle measurement and positioning are: △x=157.6846m, △y=102.6164m, △z=139.4713m. The error results are as follows: Figure 2 As shown, (a) is the XY axis position error ellipse, (b) is the YZ axis position error ellipse, and (c) is the ZX axis position error ellipse.

[0180] Step 5.2: When the detector receives the angle measurement and distance measurement signals, the linearized observation equation is obtained by using equations (11) and (15), three angle measurement information and one distance measurement information are selected, and the combined navigation positioning can be realized by solving the equation group;

[0181] make Among them, Z' is the angle measurement and ranging observation vector, and H' is the corresponding angle measurement and ranging observation matrix.

[0182] The combined observation equations are written as

[0183] Z'=H'·δr (28)

[0184] The prior estimate of the detector position r can be obtained by using the detector's orbit prediction or trajectory recursion + , use formula (4) to determine the predicted value A of the angle measurement ij + ,use Determine the predicted value R of the range i + , calculate the observation matrix H', and obtain the observation matrix Z' by subtracting the predicted value from the observed value;

[0185] The current position of the detector can be obtained by solving equation (28) using the least squares method:

[0186] r=r + +(H' T H') -1 H' T Z' (29)

[0187] In order to ensure that the simulation results are not interfered by the noise random numbers, the simulation with ranging information was performed 100 times of Monte Carlo shooting.

[0188] The average positioning errors of the three axes of the angle and distance measurement combined positioning are: △x=16.5654m, △y=16.8631m, △z=18.9542m. The error results are as follows: Figure 3 As shown, (a) is the XY axis position error ellipse, (b) is the YZ axis position error ellipse, and (c) is the ZX axis position error ellipse.

[0189] At this point, the positioning of the probe on the celestial body surface is completed.

[0190] The specific description above further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. Optimization selection method of optical navigation information for deep space probes, Features: The following steps are included: Step 1: Use sensors to obtain reference celestial body information and construct angle measurement observation equations; Step 2: Determine whether the detector can receive available ranging information, if yes, proceed to step 3, if not, proceed to step 4; Step 3: Construct the ranging observation equation based on the available ranging information obtained in step 2; Step 4: According to the angle measurement observation equation of step 1 and the distance measurement observation equation of step 3, construct an evaluation function for the angle measurement information; optimize the angle measurement information according to the evaluation function to obtain the optimal angle measurement combination, that is, to achieve the optimal selection of optical navigation information for deep space probes.

2. The method for optimizing the selection of optical navigation information of a deep space probe according to claim 1, Features: The method further includes step 5: using the optimal angle measurement combination solved in step 4, obtaining and solving the positioning equation for detector navigation.

3. The method for optimizing the selection of optical navigation information of a deep space probe according to claim 1 or 2, Features: The specific implementation method of step 1 is: The pixel coordinates p of the i-th reference celestial body from the camera's perspective i , Image line l i The coordinates are expressed as follows: Among them, c mn ,m=1,2,3n=1,2,3 is the coordinate transformation matrix C from the J2000.0 heliocentric ecliptic coordinate system to the detector body coordinate system cm The corresponding elements in the graph are x, y, and z, which are the position coordinates of the detector in the J2000.0 heliocentric ecliptic coordinate system. i ,y i , z i is the position coordinate of the reference celestial body in the J2000.0 heliocentric ecliptic coordinate system, and f is the focal length of the camera; The observation angle obtained by the camera observing any two reference celestial bodies i and j is A ij , according to the geometric relationship, we have Among them, r i With r j is the reference celestial coordinate of the detector system, r i With r j r i With r j The scalar size of The pixel and image line coordinates of the optical information obtained by equations (1) and (2) represent the observation angle, and we have: Using the vectors in the J2000.0 heliocentric ecliptic coordinate system to geometrically describe the observation angle, we have (ρ-ρ i )(ρ-ρ j )=|ρ-ρ i ||ρ-ρ j |cosA ij (5) Where ρ is the detector position vector, ρ i , ρ j is the reference celestial position vector; Linearizing equation (5) yields: Among them, δA is the angle deviation, δr is the position deviation; Let the unit vector of the line of sight from the detector to the target celestial body be Then formula (6) is transformed into Define auxiliary vector n ij With n ji Substituting formula (8) into formula (7), we have To further simplify the angle deviation δA ij The relationship between the position deviation δr is: The linearized observation equation is δA ij =h ij ·δr (11) The observation angle information of any two reference celestial bodies i and j is obtained by equation (4). According to the three pairs of information δA i1j1 , δA i2j2 and δA i3j3 That is, r is solved by equation (11); the observation angles used are: celestial body i 1 With celestial body 1 The angle between the celestial bodies 2 With celestial body 2 The angle between the celestial bodies 3 With celestial body 3 The angle between.

4. The method for optimizing the selection of optical navigation information of a deep space probe as claimed in claim 3, Features: The specific implementation method of step 3 is: At a certain moment, the detector G measured the i The pseudorange information of is obtained as follows: Among them, d k The navigation star S k Pseudorange information relative to detector G, v k The noise error generated by the observation includes the error of the detector signal receiving payload and the noise generated by the signal transmission during the process; The nominal position coordinates of the detector relative to the J2000.0 heliocentric ecliptic coordinate system are (x 0 ,y 0 ,z 0 ), Taylor expansion of equation (12) near the nominal position of the detector is obtained: Among them, R k For the probe and the navigation star S k The real distance Define auxiliary amount Let h s =[l k m k n k ], δd k =d k -R k , the linearized ranging observation equation is: δd k =h s ·δr (15)。 5. The method for optimizing the selection of optical navigation information of a deep space probe as claimed in claim 4, Features: The specific implementation method of step 4 is: Step 4.1, when there is no ranging information, directly optimize the obtained observation angle information; Using Fisher information matrix to measure the observability of optical angle measurement information in, represents the state transfer matrix, which is h in formula (11) ij ; σ i is the noise standard deviation. The measurement error variance and the trace of the Fisher information matrix satisfy Among them, λ 1 ,λ 2 ,λ 3 is the eigenvalue of the Fisher information matrix; The eigenvalue of the Fisher information matrix is ​​used as an evaluation index to measure the overall observability of a set of angle measurement information. σ 2 =tr(F) (18) When there is no distance information, the evaluation function of angle information is σ = σ 2 ; By obtaining a set of optical angle measurement information with the largest σ value, the optimal selection of the observation angle can be achieved; Step 4.2: When there is ranging information, since the positioning accuracy of ranging is much higher than that of optical angle measurement, the observation angle can be optimized according to the high-precision ranging information, and then the combined positioning method can be used to realize the positioning of the detector; The relative observability of the ranging information in the three axes is expressed by the normalized relative coordinate difference between the navigation star and the detector: Among them, OBS orbit is the relative coordinate difference between the navigation star and the detector, NOBS orbit is the normalized OBS orbit ; Calculate the inverse matrix F of the Fisher information matrix -1 The square root of the diagonal elements of , we can get the lower bound of the estimated error of each axis; using f 11 ,f 22 ,f 33 Indicates F -1 The square root of the diagonal elements of f, normalized 11 ,f 22 ,f 33 The reciprocal of indicates the relative observability of the three axes. Among them, OBS rover NOBS is the vector composed of the inverse square roots of the diagonal elements of the inverse matrix of the optical information Fisher information matrix. rover is the normalized OBS rover ; The relative observability of equation (18) and equation (22) is weighted and added to obtain a set of compensation conditions of the three axes of the angle measurement information for the distance measurement information: NOBS=NOBS orbit +κNOBS rover (23) Among them, κ is the relative observability weighting coefficient; Establish the evaluation index σ of the relative observability of angle measurement information 1 Where min(NOBS) and max(NOBS) represent the minimum and maximum values ​​of the vector NOBS respectively; σ 1 Satisfy 0≤σ 1 ≤1, the larger the value, the more likely the selected angle measurement combination is to ensure that all three axes have a higher degree of observability; Combining equation (18) and equation (24), the evaluation function of angle measurement information when distance measurement information is known is established: s = as 1 +bs 2 (25) Among them, α is the relative observability weighting coefficient, and β is the overall observability weighting coefficient. By obtaining a set of optical angle measurement information with the largest σ value, the optimization of the observation angle is achieved.

6. The method for optimizing the selection of optical navigation information of a deep space probe as claimed in claim 5, Features: The specific implementation method of step 5 is: Step 5.1: If there is no available navigation star, use the linearized angle measurement observation equation of formula (11), select three angle measurement information, and solve the equation group to achieve angle measurement positioning; make Among them, Z is the angle measurement observation vector, and H is the corresponding angle measurement observation matrix; The angle measurement observation equations are written as Z=H·δr (26) Among them, δr is the position offset, which includes the three-axis components Δx, Δy and Δz, and δr = [Δx, Δy, Δz]; Step 5.2, when the detector receives the angle measurement and distance measurement signals, the linearized observation equation is obtained by using equations (11) and (15), three angle measurement information and one distance measurement information are selected, and the combined navigation positioning is realized by solving the equation group; make Among them, Z' is the angle measurement and ranging observation vector, and H' is the corresponding angle measurement and ranging observation matrix. The combined observation equations are written as Z'=H'·δr (27) The prior estimate of the detector position r is obtained by using the detector's orbit prediction or trajectory recursion + , use formula (4) to determine the predicted value A of the angle measurement ij + ,use Determine the predicted value R of the range i + ,(x k ,y k ,z k ) is the position coordinate of the navigation star relative to the J2000.0 heliocentric ecliptic coordinate system at this time, (x + ,y + ,z + ) is r + , calculate the observation matrix H or H', and obtain the observation matrix Z or Z' by subtracting the predicted value from the observed value; The current position of the detector can be obtained by solving equation (26) or equation (27) using the least squares method. r=r + +(H T H) -1 H T Z (28)。