Minimum radar admissible domain construction method suitable for radar observation data
By building the smallest radar allowable domain in radar observation data, the problem of the inability to effectively utilize radar observation data is solved, accurate determination of orbital state and efficient calculation of arc segment correlation are achieved, and space situational awareness capabilities are improved.
Patent Information
- Application Number
- CN202411139904.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-05
AI Technical Summary
The existing minimum allowable domain technology is only suitable for optical observation data, and cannot effectively utilize radar observation data, resulting in low efficiency in orbital information utilization of radar observation data, high computational complexity, and requires prior information, which cannot meet the subsequent arc segment association. Technical requirements.
Using a minimum radar allowable domain construction method suitable for radar observation data, the minimum allowable domain is constructed on the plane of the right to change rate and declination rate, and converting azimuth and pitch angles into right to right to right and declination through the transformation matrix. Using the Levenberg-Marquardt algorithm and Jacobian matrix to calculate the accuracy information and allowable domain uncertainty error, the minimum radar allowable domain is built, the allowable domain range is narrowed and the certainty of orbital state is improved.
Without prior information, the allowable domain range of radar observation data is greatly reduced, the certainty of orbital state and the accuracy and calculation efficiency of arc segment association are improved, calculation costs and time are saved, and the calculation is made up for Minimum allowable domain technology is only suitable for defects in optical observation data.
Smart Images

Figure CN118896589A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of satellite orbit control technology in the field of high-end equipment manufacturing, and relates to a method for constructing a minimum radar allowable domain suitable for radar observation data. Background Art
[0002] As of August 15, 2024, according to data published by the Space-Track website, the number of observable space targets is nearly 46,100, including 10,300 spacecraft and 35,800 rocket bodies and debris. In near-Earth space, these targets are widely distributed in the space range of 300-36,000 km, most of which are located in the LEO region, accounting for about 80% of the total, while HEO, GEO and MEO account for 11%, 6% and 1.8% of the total, respectively. The sharp increase in the number of space targets, especially LEO targets, has put forward higher and higher requirements for space situational awareness capabilities.
[0003] One of the components of space situational awareness capability is the ability to process observation data. Different space observation methods, including radar observation and optical observation, can obtain different observation data. Optical observation methods can only obtain angular data such as right ascension and declination of space targets, and the lack of distance information often results in large uncertainties in the determination of short arc segments of angular measurement. Radar observation methods can obtain multi-dimensional information including altitude angle, azimuth angle, distance, and distance change rate, and with its all-weather and all-day working mode, it plays a vital role in the space monitoring mission of LEO targets. Therefore, improving the processing capability of radar observation data is of great practical significance.
[0004] The admissible domain method is a commonly used method for determining the initial orbit of a space target. It estimates the complete orbital state by limiting the unobservable orbital parameters by utilizing the energy constraints of orbital dynamics. Since the admissible domain method was proposed by Italian scholar Milani in 2004, it has been gradually improved to form constrained admissible domain, probability admissible domain and minimum admissible domain. The first two can be applied to radar observation data and optical observation data, while the minimum admissible domain method with the best improvement effect can only be applied to optical observation data. Reference patent (CN115859013A) is a method and system for generating the minimum admissible threshold for optical observation arc segments.
[0005] One of the main technical difficulties in space situational awareness is the inability to accurately estimate the full dynamic state of the target based on single arc segment data. This is because the target orbit information contained in the single arc segment observation data is very limited, so improving the utilization efficiency of orbit information is a beneficial way to fundamentally improve situational awareness capabilities. The optical minimum allowable domain method efficiently utilizes optical observation data. By deriving the uncertainty propagation theory, it can greatly reduce the range of the allowable domain without the need for orbital prior information, thereby greatly improving the observation arc segment association efficiency and accuracy. However, the existing minimum allowable domain theory is only applicable to optical observation data, which contains right ascension and declination. The distance and distance change rate are used as parameters to be estimated, and the possible orbital state that can describe the space target is constructed on the (distance, distance change rate) plane. However, this theory cannot be applied to radar observation data.
[0006] Radar observation data include altitude angle, azimuth angle, distance and distance change rate, and it is necessary to construct an admissible domain on the (right ascension change rate, declination change rate) plane. The admissible domain theory applicable to radar observation data includes constrained admissible domain and probabilistic admissible domain. These two theoretical methods sometimes need to rely on prior information, and the degree of reduction of the admissible domain is limited, which cannot meet the technical requirements of subsequent arc segment association. Summary of the invention
[0007] In view of the fact that the original radar allowable domain is too large, the probability density distribution and arc segment association calculation based on it is too costly, and the satellite prior information is required, and the calculation complexity is too large; and the technical problem that the current minimum allowable domain technology is only applicable to optical observation data, the present invention provides a minimum radar allowable domain construction method applicable to radar observation data. The method constructs the minimum allowable domain on the (right ascension change rate, declination change rate) plane for radar observation data, minimizes the scope of the original radar allowable domain, and limits the possible orbital state of the space target to a very small area, so as to improve the accuracy and calculation efficiency of subsequent arc segment association. The present invention utilizes radar observation data to greatly reduce the scope of the original allowable domain on the two-dimensional plane without any prior information of any other satellites. It makes up for the defect that the current minimum allowable domain technology is only applicable to optical observation data, and extends the minimum allowable domain theory to radar observation data.
[0008] The purpose of the present invention is specifically achieved through the following technical solutions:
[0009] The present invention discloses a method for constructing a minimum radar permissible domain applicable to radar observation data, the method comprising:
[0010] Step 1: Use the conversion matrix to convert the azimuth and elevation angles in the actual radar observation data received from the station coordinate system to the geocentric inertial coordinate system to obtain the right ascension and declination;
[0011] Step 2: Construct arc attributes from the four parameters of right ascension, declination, distance and distance change rate in actual radar observation data, and obtain the orbit elements of the allowed domain from the constructed arc attributes; based on the characteristic that the orbit energy of space targets in near-Earth space is less than 0, the orbit energy is represented by the orbit elements of the allowed domain, and the original radar allowed domain is constructed on the allowed domain plane composed of the right ascension change rate and the declination change rate;
[0012] Step 3: According to the observation equation, the orbital elements of the original radar allowed domain are converted into orbital states in the Cartesian coordinate system; and then the orbital states are mapped into theoretical radar observation data;
[0013] Step 4: Construct the first loss function based on the actual radar observation data and observation error, and use the Levenberg-Marquardt algorithm to find the least square solution of the first loss function in the original radar admissible domain;
[0014] Step 5: Calculate the accuracy information of the four parameters of right ascension, declination, distance and distance change rate in the actual radar observation data through the normal equation;
[0015] Step 6: Calculate the allowable domain uncertainty error of the accuracy information through the Jacobian matrix; combine the observation error to calculate the minimum allowable domain error of the actual radar observation data;
[0016] Step 7: Construct a loss function through the minimum allowable domain error of the actual radar observation data and the theoretical radar observation data, sample around the least squares solution of the original radar allowable domain, and use the sampling point when the loss function value is equal to 3 as the boundary of the minimum radar allowable domain to construct the minimum radar allowable domain.
[0017] The transformation matrix in step 1 is:
[0018]
[0019] in, is the actual arc segment attribute, that is, the actual radar observation data in the geocentric inertial coordinate system; S is the station longitude, Ω G It is the angle between the geocentric inertial coordinate system and the X-axis of the geocentric coordinate system, which can be calculated by looking up the astronomical almanac; is the latitude of the measuring station; B rad is the actual radar observation data in the station coordinate system; M 3 [.] is the third elementary transformation matrix; M 2 [.] is the second elementary transformation matrix; where,
[0020]
[0021] Among them, M2 [θ] is the second elementary transformation matrix about angle θ; M 3 [θ] is the third elementary transformation matrix about angle θ; θ is the rotation angle.
[0022] In step 2, the method for constructing arc attributes is:
[0023]
[0024] Among them, B is the constructed arc attribute; ρ is the distance, is the rate of change of distance, α is the right ascension, and δ is the declination; is a positive real number; is the set of all real numbers;
[0025] The number of orbital elements in the allowed domain obtained from the constructed arc segment attributes is:
[0026]
[0027] in, is the number of orbital elements in the allowed domain; ρ is the distance, is the distance change rate, α is the right ascension, δ is the declination, is the rate of change of right ascension; is the rate of change of declination.
[0028] In step 2, the method for constructing the original radar admissible domain on the admissible domain plane composed of the right ascension change rate and the declination change rate is:
[0029]
[0030] in, is the rate of change of right ascension; is the rate of change of declination; z 1 is the first coefficient; z 2 is the second coefficient; z 3 is the third coefficient:
[0031]
[0032] Among them, Z is the coefficient matrix, which is a real symmetric matrix; 11 is the eleventh coefficient; z 12 is the twelfth coefficient; z 13 is the thirteenth coefficient; z 21 is the twenty-first coefficient; z 22 is the 22nd coefficient; z 23 is the twenty-third coefficient; z 31 is the thirty-first coefficient; z 32 is the thirty-second coefficient; z 33 is the thirty-third coefficient;
[0033]
[0034] Where ρ is the distance; δ is the declination; ω 3,α is the seventh sub-coefficient; ω 3,δ is the eighth sub-coefficient; is the distance change rate; ω 1 is the second sub-coefficient; ω 4 is the fifth sub-coefficient; μ E is the earth's gravitational constant; α is the right ascension; S(ρ) is the coefficient polynomial:
[0035] Where S(ρ) = ρ 2 +ω 5 ρ+ω 0 ;
[0036] Among them, ω 0 is the first sub-coefficient; ω 5 is the sixth sub-coefficient:
[0037] ω 0 =||R|| 2 ;
[0038]
[0039] ω 5 =2R·L;
[0040]
[0041] Where R is the position vector of the observation platform; is the velocity vector of the observation platform; L is the unit vector of the satellite relative to the observation platform; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; L δ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination.
[0042] The observation equation in step 3 is:
[0043]
[0044] Among them, the track state r is the satellite position vector, is the velocity vector of the satellite; R is the position vector of the observation platform, is the velocity vector of the observation platform; ρ is the distance, is the distance change rate; L is the unit vector of the satellite relative to the observation platform; is the rate of change of right ascension; is the rate of change of declination; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; Lδ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination;
[0045] The mapping relationship between orbital state and theoretical radar observation data is:
[0046]
[0047] in, is the mapping of distance; is the mapping of right ascension; is the mapping of declination; I 3ρ (x) is the mapping of orbital state to distance; I 3α (x) is the mapping of orbital state to right ascension; 3δ (x) is the mapping of orbital state to declination; x is the orbital state, in,
[0048]
[0049] R is the position vector of the observation platform; r is the position vector of the satellite; is the satellite's velocity vector; ρ is the distance, α is the right ascension, and δ is the declination; ρ x is the component of the distance on the x-axis; ρ y is the component of the distance on the y-axis; ρ z is the component of the distance on the z-axis.
[0050] In step 4, the least squares solution of the first loss function in the original radar admissible domain is obtained using the Levenberg-Marquardt algorithm:
[0051] Use the Levenberg-Marquardt algorithm to find the first loss function The minimum value of is the least square solution of the first loss function in the original radar admissible domain; wherein the first loss function is constructed based on the actual radar observation data and radar observation error for:
[0052]
[0053] Wherein, m is the total number of radar observation data; i is the sequence number of radar observation data;
[0054] is the actual radar observation distance; is the radar observation distance to be estimated; is the distance observation error;
[0055] is the actual radar observation right ascension; is the radar observation right ascension to be estimated; is the right ascension observation error; is the actual radar observation declination; is the radar observation declination to be estimated; σ 2 is the declination observation error.
[0056] In step 5, the normal equation is:
[0057] P=Λ -1 ;
[0058]
[0059] Where P = Λ -1 is the covariance matrix of the measured quantity, including the accuracy information of right ascension, declination, distance and distance change rate; Λ is the information matrix; is the observation error; n is the number of observation points; T is the time span of the observation arc;
[0060] Through the normal equation, the accuracy information of the four parameters of right ascension, declination, distance and distance change rate are calculated as follows:
[0061]
[0062] in, are the accuracy information of distance, distance change rate, right ascension and declination, σ ρ is the radar ranging accuracy, σ is the radar angle measurement accuracy, T is the time span of the observation arc, and N is the number of observation points.
[0063] In step 6, the allowable domain uncertainty error of the accuracy information is calculated by the Jacobian matrix:
[0064]
[0065]
[0066] in, is the allowable uncertainty error of the accuracy information of right ascension, The theoretical radar observation data for the accuracy information of right ascension, The real radar observation data of the accuracy information of right ascension; is the Jacobian matrix of right ascension propagation; is the real arc attribute; is the actual arc attribute;
[0067] is the allowable domain uncertainty error of the distance accuracy information, The theoretical radar observation data for distance accuracy information, Real radar observation data for distance accuracy information; is the Jacobian matrix of distance propagation;
[0068] is the allowable uncertainty error of the accuracy information of declination, Theoretical radar observation data for the accuracy information of declination, Real radar observation data for the accuracy information of declination; is the Jacobian matrix of declination propagation;
[0069]
[0070] in, is the propagation function of right ascension; is the propagation function of declination; is the propagation function of distance; B is the arc attribute; is the actual arc attribute; ρ i is the distance at the i-th observation point, α i is the declination of the i-th observation point at the moment, δ i is the declination of the i-th observation point at the moment, x i is the orbital state at the i-th observation point; x 0 is the orbital state at the initial moment;
[0071] In step 6, the minimum allowable domain error of the actual radar observation data is:
[0072]
[0073] in, is the minimum allowable domain error of the distance; is the actual radar observation distance of the ith object; is the i-th theoretical radar observation distance; represents normal distribution; is the distance observation error; is the Jacobian matrix of distance propagation; Σ B is the uncertainty matrix;
[0074] is the minimum allowable domain error of right ascension; is the right ascension of the ith actual radar observation; is the right ascension of the ith theoretical radar observation; is the right ascension observation error; is the Jacobian matrix of right ascension propagation;
[0075] is the minimum allowable domain error of declination; is the i-th actual radar observation declination; is the i-th theoretical radar observation declination; σ2 is the angle observation error; is the Jacobian matrix of declination propagation;
[0076]
[0077] in, They are the accuracy information of distance, distance change rate, right ascension and declination respectively.
[0078] In step 7, the loss function constructed for:
[0079]
[0080] The loss function is used to calculate the loss function value of each sampling point in the original radar permissible domain, and the threshold β is set to ≥ 3;
[0081] Eliminate the loss function value The sampling points that meet The sampling points of are included in the minimum radar allowable domain as the boundary of the minimum radar allowable domain, and the minimum radar allowable domain is constructed;
[0082] Wherein, m is the total number of radar observation data; i is the sequence number of radar observation data;
[0083] is the actual radar observation distance; is the theoretical radar observation distance; is the distance observation error; is the Jacobian matrix of the distance propagation corresponding to the least squares solution orbit; Σ B is the uncertainty matrix;
[0084] is the actual radar observation right ascension; is the theoretical radar observation right ascension; is the right ascension observation error; Φ αiLS is the Jacobian matrix of the right ascension propagation corresponding to the least squares solution orbit;
[0085] is the actual radar observation declination; is the theoretical radar observation declination; σ 2 is the declination observation error; is the Jacobian matrix of the declination propagation corresponding to the least squares solution orbit.
[0086] The beneficial effects of the present invention are:
[0087] 1. The present invention converts the azimuth and elevation angles in the actual radar observation data received from the station coordinate system to the geocentric inertial coordinate system through a conversion matrix to obtain the right ascension and declination, which can facilitate subsequent calculations;
[0088] 2. Construct the original radar admissible domain on the admissible domain plane formed by the right ascension change rate and the declination change rate, and use the Levenberg-Marquardt algorithm to find the sampling point that best matches the observation data in the original radar admissible domain. The least squares solution of the sampling points in the original radar admissible domain is obtained to describe the orbit that best matches the observation data. The corresponding orbital state is the initial orbit determination result of this observation arc segment. This is equivalent to an initial orbit determination method based on the radar admissible domain, which can provide a more accurate initial value for subsequent precise orbit determination.
[0089] 3. According to the observation equation, the orbital elements of the original radar allowed domain are converted into the orbital state in the Cartesian coordinate system; the orbital state is then mapped into the theoretical radar observation data. The conversion between the orbital elements of the allowed domain, the Cartesian orbital state and the theoretical radar observation data is realized.
[0090] 4. From a mathematical perspective, the present invention calculates the accuracy information of the four parameters of right ascension, declination, distance and distance change rate in radar observation data through normal equations; calculates the allowable domain uncertainty error of the accuracy information through the Jacobian matrix; combines the radar observation error to calculate the minimum allowable domain error of the actual radar observation data, and analytically expresses the uncertainty of the actual observation data. This lays a theoretical foundation for the uncertainty analysis and arc segment association of the subsequent initial orbit determination results.
[0091] 5. The loss function is constructed by using the minimum allowable domain error of actual radar observation data and theoretical radar observation data. The sampling points (right ascension change rate, declination change rate) when the loss function value is equal to 3 are used as the boundary of the minimum radar allowable domain, and the minimum radar allowable domain is obtained. The 3σ criterion of normal distribution is introduced to provide theoretical support for the determination of the boundary, which is also more in line with the actual situation.
[0092] 6. The present invention utilizes actual radar observation data to greatly reduce the scope of the original allowable domain on the two-dimensional plane without any other prior information. The efficient use of actual radar observation data greatly reduces the scope of the original radar allowable domain, limits the possible orbital states to a very small area, and increases the accuracy and efficiency of the initial orbit determination; at the same time, it improves the accuracy and calculation efficiency of subsequent arc segment associations, greatly saving the calculation cost and time of subsequent arc segment associations. It makes up for the defect that the current minimum allowable domain technology is only applicable to optical observation data, and extends the minimum allowable domain theory to radar observation data. BRIEF DESCRIPTION OF THE DRAWINGS
[0093] The present invention is further described in detail below based on the accompanying drawings and embodiments.
[0094] Figure 1 It is a schematic diagram of the original radar allowable domain.
[0095] Figure 2 This is a schematic diagram of the location of the least squares solution in the original radar admissible domain.
[0096] Figure 3 This is a schematic diagram of the minimum radar allowable area.
[0097] Figure 4 Schematic diagram of the location of the true value and the least squares solution in the minimum radar admissible domain. DETAILED DESCRIPTION
[0098] Embodiment 1
[0099] Embodiment 1 of the present invention provides a method for constructing a minimum radar allowable domain applicable to radar observation data, the method comprising:
[0100] Step 1: Use the conversion matrix to convert the azimuth and elevation angles in the actual radar observation data received from the station coordinate system to the geocentric inertial coordinate system to obtain the right ascension and declination;
[0101] Step 2: Construct arc attributes from the four parameters of right ascension, declination, distance and distance change rate in actual radar observation data, and obtain the orbit elements of the allowed domain from the constructed arc attributes; based on the characteristic that the orbit energy of space targets in near-Earth space is less than 0, the orbit energy is represented by the orbit elements of the allowed domain, and the original radar allowed domain is constructed on the allowed domain plane composed of the right ascension change rate and the declination change rate;
[0102] Step 3: According to the observation equation, the orbital elements of the original radar allowed domain are converted into orbital states in the Cartesian coordinate system; and then the orbital states are mapped into theoretical radar observation data;
[0103] Step 4: Construct the first loss function based on the actual radar observation data and observation error, and use the Levenberg-Marquardt algorithm to find the least square solution of the first loss function in the original radar admissible domain;
[0104] Step 5: Calculate the accuracy information of the four parameters of right ascension, declination, distance and distance change rate in the actual radar observation data through the normal equation;
[0105] Step 6: Calculate the allowable domain uncertainty error of the accuracy information through the Jacobian matrix; combine the observation error to calculate the minimum allowable domain error of the actual radar observation data;
[0106] Step 7: Construct a loss function through the minimum allowable domain error of the actual radar observation data and the theoretical radar observation data, sample around the least squares solution of the original radar allowable domain, and use the sampling point when the loss function value is equal to 3 as the boundary of the minimum radar allowable domain to construct the minimum radar allowable domain.
[0107] The transformation matrix in step 1 is:
[0108]
[0109] in, is the actual arc segment attribute, that is, the actual radar observation data in the geocentric inertial coordinate system; S is the station longitude, Ω G It is the angle between the geocentric inertial coordinate system and the X-axis of the geocentric coordinate system, which can be calculated by looking up the astronomical almanac; is the latitude of the measuring station; B rad is the actual radar observation data in the station coordinate system; M 3 [.] is the third elementary transformation matrix; M 2 [.] is the second elementary transformation matrix; where,
[0110]
[0111] Among them, M 2 [θ] is the second elementary transformation matrix about angle θ; M 3 [θ] is the third elementary transformation matrix about angle θ; θ is the rotation angle.
[0112] In step 2, the method for constructing arc attributes is:
[0113]
[0114] Among them, B is the constructed arc attribute; ρ is the distance, is the rate of change of distance, α is the right ascension, and δ is the declination; is a positive real number; is the set of all real numbers;
[0115] The number of orbital elements in the allowed domain obtained from the constructed arc segment attributes is:
[0116]
[0117] in, is the number of orbital elements in the allowed domain; ρ is the distance, is the distance change rate, α is the right ascension, δ is the declination, is the rate of change of right ascension; is the rate of change of declination.
[0118] In this step, the actual radar observation data is compressed into "arc segment attributes" (Attributable), retaining only the distance, distance change rate and angle information at a certain moment. The four parameters of the arc segment attributes: distance, distance change rate, right ascension and declination almost contain all the valid information of the actual radar observation arc segment.
[0119] In step 2, the method for constructing the original radar admissible domain on the admissible domain plane composed of the right ascension change rate and the declination change rate is:
[0120]
[0121] It can be seen that the original radar admissible domain under energy constraints is It is elliptical in shape on a flat surface.
[0122] in, is the rate of change of right ascension; is the rate of change of declination; z 1 is the first coefficient; z 2 is the second coefficient; z 3 is the third coefficient:
[0123]
[0124] Among them, Z is the coefficient matrix, which is a real symmetric matrix; 11 is the eleventh coefficient; z 12 is the twelfth coefficient; z 13 is the thirteenth coefficient; z 21 is the twenty-first coefficient; z 22 is the 22nd coefficient; z 23 is the twenty-third coefficient; z 31 is the thirty-first coefficient; z 32 is the thirty-second coefficient; z 33 is the thirty-third coefficient;
[0125]
[0126] Where ρ is the distance; δ is the declination; ω 3,α is the seventh sub-coefficient; ω 3,δ is the eighth sub-coefficient; is the distance change rate; ω 1 is the second sub-coefficient; ω 4 is the fifth sub-coefficient; μ E is the earth's gravitational constant; α is the right ascension; S(ρ) is the coefficient polynomial:
[0127] Where S(ρ) = ρ 2 +ω 5 ρ+ω 0 ;
[0128] Among them, ω 0 is the first sub-coefficient; ω 5 is the sixth sub-coefficient:
[0129] ω 0 =||R|| 2 ;
[0130]
[0131] ω 5 =2R·L;
[0132]
[0133] Where R is the position vector of the observation platform; is the velocity vector of the observation platform; L is the unit vector of the satellite relative to the observation platform; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; L δ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination.
[0134] The observation equation in step 3 is:
[0135]
[0136] Among them, the track state r is the satellite position vector, is the velocity vector of the satellite; R is the position vector of the observation platform, is the velocity vector of the observation platform; ρ is the distance, is the distance change rate; L is the unit vector of the satellite relative to the observation platform; is the rate of change of right ascension; is the rate of change of declination; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; L δ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination;
[0137] The mapping relationship between orbital state and theoretical radar observation data is:
[0138]
[0139] in, is the mapping of distance; is the mapping of right ascension; is the mapping of declination; I 3ρ (x) is the mapping of orbital state to distance; I 3α (x) is the mapping of orbital state to right ascension; 3δ(x) is the mapping of orbital state to declination; x is the orbital state, in,
[0140]
[0141] R is the position vector of the observation platform; r is the position vector of the satellite; is the satellite's velocity vector; ρ is the distance, α is the right ascension, and δ is the declination; ρ x is the component of the distance on the x-axis; ρ y is the component of the distance on the y-axis; ρ z is the component of the distance on the z-axis.
[0142] In step 4, the least squares solution of the first loss function in the original radar admissible domain is obtained using the Levenberg-Marquardt algorithm:
[0143] In the admissible domain, we can find an orbit that best fits the observed data. According to its corresponding orbit, the radar observation data to be estimated can be calculated The weighted square sum of the theoretical radar observation sequence and the actual radar sensing sequence can characterize the estimation accuracy of the orbit state and construct the first loss function
[0144] Use the Levenberg-Marquardt algorithm to find the first loss function The minimum value of is the least squares solution of the sampling points in the original radar admissible domain; where,
[0145]
[0146] in, is the first loss function; m is the total number of radar observation data; i is the sequence number of radar observation data;
[0147] is the actual radar observation distance; is the radar observation distance to be estimated; is the distance observation error;
[0148] is the actual radar observation right ascension; is the radar observation right ascension to be estimated; is the right ascension observation error; is the actual radar observation declination; is the radar observation declination to be estimated; σ 2 is the declination observation error.
[0149] The least squares solution represents the optimal orbit state estimate within the original radar tolerance domain. It should be noted that due to the existence of observation errors, the true right ascension change rate and declination change rate are It is not a least squares solution in general. However, relative to the range of the admissible domain, the two are relatively close.
[0150] In step 5, the normal equation is:
[0151] P=Λ -1 ;
[0152]
[0153] Where P = Λ -1 is the covariance matrix of the measured quantity, including the accuracy information of right ascension, declination, distance and distance change rate; Λ is the information matrix; is the observation error; n is the number of observation points; T is the time span of the observation arc;
[0154] Through the normal equation, the uncertainty of the actual observation data is analyzed. The additional precision hidden information can be used to reduce the uncertainty of the measured quantity, derive the precision information of the angle observation value (right ascension and declination) and the distance observation value (distance and distance change rate); calculate the precision information of the four parameters of right ascension, declination, and distance and distance change rate in the radar observation data:
[0155]
[0156] in, are the accuracy information of distance, distance change rate, right ascension and declination, σ ρ is the radar ranging accuracy, σ is the radar angle measurement accuracy, T is the time span of the observation arc, and N is the number of observation points.
[0157] In step 6, the allowable domain uncertainty error of the accuracy information is calculated by the Jacobian matrix:
[0158]
[0159] in, is the allowable uncertainty error of the accuracy information of right ascension, The theoretical radar observation data for the accuracy information of right ascension, The real radar observation data of the accuracy information of right ascension; is the Jacobian matrix of right ascension propagation; is the real arc attribute, that is, the real state of the satellite in the sky; is the actual arc attribute, and is the observed data of the real arc attribute;
[0160] is the allowable domain uncertainty error of the distance accuracy information, The theoretical radar observation data for distance accuracy information, Real radar observation data for distance accuracy information; is the Jacobian matrix of distance propagation;
[0161] is the allowable uncertainty error of the accuracy information of declination, Theoretical radar observation data for the accuracy information of declination, Real radar observation data for the accuracy information of declination; is the Jacobian matrix of declination propagation;
[0162]
[0163] in, is the propagation function of right ascension; is the propagation function of declination; is the propagation function of distance; B is the arc attribute; is the actual arc attribute; ρ i is the distance at the i-th observation point, α i is the declination of the i-th observation point at the moment, δ i is the declination of the i-th observation point at the moment, x i is the orbital state at the i-th observation point; x 0 is the orbital state at the initial moment;
[0164] In step 6, the minimum allowable domain error of the actual radar observation data is:
[0165]
[0166] in, is the minimum allowable domain error of the distance; is the actual radar observation distance of the ith object; is the i-th theoretical radar observation distance; represents normal distribution; is the distance observation error; is the Jacobian matrix of distance propagation; Σ B is the uncertainty matrix;
[0167] is the minimum allowable domain error of right ascension; is the right ascension of the ith actual radar observation; is the right ascension of the ith theoretical radar observation; is the right ascension observation error; is the Jacobian matrix of right ascension propagation;
[0168] is the minimum allowable domain error of declination; is the i-th actual radar observation declination; is the i-th theoretical radar observation declination; σ 2 is the angle observation error; is the Jacobian matrix of declination propagation;
[0169] The above formula shows that the uncertainty of the allowable domain has an amplifying effect on the angle error. The theoretical observation errors of distance, right ascension, and declination are affected by two parts: the first part is the observation error of the sensor. σ 2 , the second part is the allowable domain uncertainty error
[0170] in,
[0171] in, They are the accuracy information of distance, distance change rate, right ascension and declination respectively.
[0172] In step 7, the minimum radar permissible region needs to contain the true right ascension change rate and declination change rate with a high probability. In practice, is unknown, but fortunately, the least squares solution can be used to approximate the true value, the matrix It can also be approximated by substituting the values of the least squares solution. The Jacobian matrices under the least squares solution are Constructed loss function for:
[0173]
[0174] Each cumulative term represents the standardized deviation of a theoretical observation from the actual observation (absolute error divided by the standard deviation). It represents the average angular deviation of the track. Referring to the 3σ criterion of normal distribution, the loss function value of each sampling point in the original radar permissible domain is calculated by the loss function, and the threshold β is set to ≥ 3; that is, the probability that the constructed minimum radar permissible domain contains the true track is greater than 99.74%; for any point in the permissible domain If the loss function If it is less than β, it is included in the minimum allowable domain; otherwise, it is considered impossible to represent the true orbit state. will be accepted with great probability; points that are far from the true orbit will be rejected due to large angular errors.
[0175] Perform uniform sampling in the admissible domain and condition each sampling point The minimum allowable domain boundary can be determined by judgment.
[0176] Eliminate the loss function value The sampling points that meet The sampling points of are included in the minimum radar allowable domain as the boundary of the minimum radar allowable domain, and the minimum radar allowable domain is constructed;
[0177] Wherein, m is the total number of radar observation data; i is the sequence number of radar observation data;
[0178] is the actual radar observation distance; is the theoretical radar observation distance; is the distance observation error; is the Jacobian matrix of the distance propagation corresponding to the least squares solution orbit; Σ B is the uncertainty matrix;
[0179] is the actual radar observation right ascension; is the theoretical radar observation right ascension; is the right ascension observation error; is the Jacobian matrix of the right ascension propagation corresponding to the least squares solution orbit;
[0180] is the actual radar observation declination; is the theoretical radar observation declination; σ 2 is the declination observation error; is the Jacobian matrix of the declination propagation corresponding to the least squares solution orbit.
[0181] Verify the instance
[0182] The embodiment of the present invention takes a simulated radar observation arc of a LEO target as an example to construct the corresponding minimum radar permissible domain. The following table shows the specific information and observation error information of the selected space target.
[0183] Table 1
[0184] Target Information Numeric Epoch time (UTCG) 20Mar 2020 15:04:37.000 a / km 6923.161730 e 0.000832 i / ° 53.149 Ω / ° 341.470 w / ° 154.344 f / ° 346.534 Observation arc duration / s 61 Angle observation error / ° 0.1 Distance observation error / m 30
[0185] The minimum radar permissible region is constructed based on the information disclosed in Table 1.
[0186] The arc segment attributes of radar observation data are constructed by four parameters: right ascension, declination, distance and distance change rate; the original radar permissible domain is constructed according to the characteristic that the orbital energy of space targets in near-Earth space is less than 0. The constructed original radar permissible domain is as follows: Figure 1 As shown. The horizontal axis represents the rate of change of right ascension, and the vertical axis represents the rate of change of declination. Figure 1It can be seen that the original admissible domain of the radar data is an ellipse, which verifies The theoretical correctness.
[0187] Sampling is performed within the original radar admissible domain, and each sampling point represents a possible orbit. An orbit that best fits the observed data is found within the original radar admissible domain, which corresponds to a set of (right ascension change rate, declination change rate), which is the least squares solution of the original radar admissible domain. The position of the least squares solution in the original radar admissible domain is as follows: Figure 2 shown.
[0188] Perform uniform sampling in the original radar admissible domain and condition each sampling point The minimum radar permissible domain boundary is determined by the judgment. In this embodiment, the value of β is 10. Figure 3 Indicates the position and size relationship of the minimum radar allowable domain in the original radar allowable domain.
[0189] Figure 4 Represents the location of the least squares solution and the true value in the minimum radar admissible region. Figure 4 In the equation, “●” indicates the position of the least squares solution and “*” indicates the position of the true value. Figure 4 It can be seen that the least square solution in the radar admissible domain is very close to the true value, and both are distributed in the constructed minimum radar admissible domain. The true value is the true value of the right ascension change rate and the declination change rate.
[0190] The beneficial effects of the embodiments of the present invention are:
[0191] 1. The present invention converts the azimuth and elevation angles in the actual radar observation data received from the station coordinate system to the geocentric inertial coordinate system through a conversion matrix to obtain the right ascension and declination, which can facilitate subsequent calculations;
[0192] 2. Construct the original radar admissible domain on the admissible domain plane formed by the right ascension change rate and the declination change rate, and use the Levenberg-Marquardt algorithm to find the sampling point that best matches the observation data in the original radar admissible domain. The least squares solution of the sampling points in the original radar admissible domain is obtained to describe the orbit that best matches the observation data. The corresponding orbital state is the initial orbit determination result of this observation arc segment. This is equivalent to an initial orbit determination method based on the radar admissible domain, which can provide a more accurate initial value for subsequent precise orbit determination.
[0193] 3. According to the observation equation, the orbital elements of the original radar allowed domain are converted into orbital states in the Cartesian coordinate system; then the orbital states are mapped to theoretical radar observation data, realizing the conversion between the orbital elements of the allowed domain, Cartesian orbital states and theoretical radar observation data.
[0194] 4. From a mathematical perspective, the present invention calculates the accuracy information of four parameters, namely right ascension, declination, distance and distance change rate in radar observation data, through normal equations; calculates the allowable domain uncertainty error of the accuracy information through the Jacobian matrix; combines the radar observation error to calculate the minimum allowable domain error of the actual radar observation data, and analytically expresses the uncertainty of the actual observation data, laying a theoretical foundation for the uncertainty analysis and arc segment association of the subsequent preliminary orbit determination results.
[0195] 5. The loss function is constructed through the minimum allowable domain error of actual radar observation data and theoretical radar observation data. The sampling points (right ascension change rate, declination change rate) when the loss function value is equal to 3 are used as the boundaries of the minimum radar allowable domain. The minimum radar allowable domain is obtained, and the 3σ criterion of normal distribution is introduced, which provides theoretical support for the determination of the boundary and is more in line with the actual situation.
[0196] 6. The present invention utilizes actual radar observation data to greatly reduce the scope of the original allowable domain on the two-dimensional plane without any other prior information. The efficient use of actual radar observation data greatly reduces the scope of the original radar allowable domain, limits the possible orbital states to a very small area, and increases the accuracy and efficiency of the initial orbit determination; at the same time, it improves the accuracy and calculation efficiency of subsequent arc segment associations, greatly saving the calculation cost and time of subsequent arc segment associations. It makes up for the defect that the current minimum allowable domain technology is only applicable to optical observation data, and extends the minimum allowable domain theory to radar observation data.
[0197] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A method for constructing a minimum radar admissible region suitable for radar observation data, characterized in that: The method includes: Step 1: Convert the azimuth and elevation angles in the actual radar observation data received from the station coordinate system to the geocentric inertial coordinate system through the conversion matrix to obtain the right ascension and declination; Step 2: construct arc attributes from the four parameters of right ascension, declination, distance and distance change rate in actual radar observation data, and obtain the orbit elements of the allowed domain from the constructed arc attributes; based on the characteristic that the orbit energy of space targets in near-Earth space is less than 0, the orbit energy is represented by the orbit elements of the allowed domain, and the original radar allowed domain is constructed on the allowed domain plane composed of the right ascension change rate and the declination change rate; Step 3: According to the observation equation, the orbital elements of the original radar allowed domain are converted into orbital states in the Cartesian coordinate system; and then the orbital states are mapped into theoretical radar observation data; Step 4: construct a first loss function based on actual radar observation data and observation errors, and use the Levenberg-Marquardt algorithm to find the least squares solution of the first loss function within the original radar admissible domain; Step 5: Calculate the accuracy information of four parameters, namely, right ascension, declination, distance and distance change rate in actual radar observation data, through normal equations; Step 6: Calculate the allowable domain uncertainty error of the accuracy information through the Jacobian matrix; combine the observation error to calculate the minimum allowable domain error of the actual radar observation data; Step 7: Construct a loss function through the minimum allowable domain error of the actual radar observation data and the theoretical radar observation data, sample around the least squares solution of the original radar allowable domain, and use the sampling point when the loss function value is equal to 3 as the boundary of the minimum radar allowable domain to construct the minimum radar allowable domain.
2. The method according to claim 1, characterized in that The transformation matrix in step 1 is: in, is the actual arc segment attribute, that is, the actual radar observation data in the geocentric inertial coordinate system; S is the station longitude, Ω G It is the angle between the geocentric inertial coordinate system and the X-axis of the geocentric coordinate system, which can be calculated by looking up the astronomical almanac; is the latitude of the measuring station; B rad is the actual radar observation data in the station coordinate system; M3[.] is the third elementary transformation matrix; M2[.] is the second elementary transformation matrix; where, Among them, M2[θ] is the second elementary transformation matrix about angle θ; M3[θ] is the third elementary transformation matrix about angle θ; θ is the rotation angle.
3. The method according to claim 1, characterized in that In step 2, the method for constructing arc attributes is: Among them, B is the constructed arc attribute; ρ is the distance, is the rate of change of distance, α is the right ascension, and δ is the declination; is a positive real number; is the set of all real numbers; The number of orbital elements in the allowed domain obtained from the constructed arc segment attributes is: in, is the number of orbital elements in the allowed domain; ρ is the distance, is the distance change rate, α is the right ascension, δ is the declination, is the rate of change of right ascension; is the rate of change of declination.
4. The method according to claim 3, characterized in that In step 2, the method for constructing the original radar admissible domain on the admissible domain plane composed of the right ascension change rate and the declination change rate is: in, is the rate of change of right ascension; is the declination change rate; z1 is the first coefficient; z2 is the second coefficient; z3 is the third coefficient: Among them, Z is the coefficient matrix, which is a real symmetric matrix; 11 is the eleventh coefficient; z 12 is the twelfth coefficient; z 13 is the thirteenth coefficient; z 21 is the twenty-first coefficient; z 22 is the 22nd coefficient; z 23 is the twenty-third coefficient; z 31 is the thirty-first coefficient; z 32 is the thirty-second coefficient; z 33 is the thirty-third coefficient; Among them, α is right ascension; ρ is distance; δ is declination; ω 3,α is the seventh sub-coefficient; ω 3,δ is the eighth sub-coefficient; is the distance change rate; ω1 is the second sub-coefficient; ω4 is the fifth sub-coefficient; μ E is the earth's gravitational constant; S(ρ) is the coefficient polynomial: Where, S(ρ)=ρ 2 +ω5ρ+ω0; Among them, ω0 is the first sub-coefficient; ω5 is the sixth sub-coefficient: ω0=||R|| 2 ; ω5=2R·L; Where R is the position vector of the observation platform; is the velocity vector of the observation platform; L is the unit vector of the satellite relative to the observation platform; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; L δ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination.
5. The method according to claim 1, characterized in that The observation equation in step 3 is: Among them, the track state r is the satellite position vector, is the velocity vector of the satellite; R is the position vector of the observation platform, is the velocity vector of the observation platform; ρ is the distance, is the distance change rate; L is the unit vector of the satellite relative to the observation platform; is the rate of change of right ascension; is the rate of change of declination; L α is the partial derivative of the satellite's unit vector relative to the observation platform with respect to right ascension; L δ is the partial derivative of the satellite's unit vector with respect to the observation platform with respect to declination; The mapping relationship between orbital state and theoretical radar observation data is: in, is the mapping of distance; is the mapping of right ascension; is the mapping of declination; I 3ρ (x) is the mapping of orbital state to distance; I 3α (x) is the mapping of orbital state to right ascension; 3δ (x) is the mapping of orbital state to declination; x is the orbital state, in, R is the position vector of the observation platform; r is the position vector of the satellite; is the satellite's velocity vector; ρ is the distance, α is the right ascension, and δ is the declination; ρ x is the component of the distance on the x-axis; ρ y is the component of the distance on the y-axis; ρ z is the component of the distance on the z-axis.
6. The method according to claim 1, characterized in that In step 4, the least squares solution of the first loss function in the original radar admissible domain is obtained using the Levenberg-Marquardt algorithm: Use the Levenberg-Marquardt algorithm to find the first loss function The minimum value of is the least square solution of the first loss function in the original radar admissible domain; wherein the first loss function is constructed based on the actual radar observation data and radar observation error for: Wherein, m is the total number of radar observation data; i is the sequence number of radar observation data; is the actual radar observation distance; is the radar observation distance to be estimated; is the distance observation error; is the actual radar observation right ascension; is the radar observation right ascension to be estimated; is the right ascension observation error; is the actual radar observation declination; is the radar observation declination to be estimated; σ 2 is the declination observation error.
7. The method according to claim 1, characterized in that In step 5, the normal equation is: P=Λ -1 ; Where P = Λ -1 is the covariance matrix of the measured quantity, including the accuracy information of right ascension, declination, distance and distance change rate; Λ is the information matrix; is the observation error; n is the number of observation points; T is the time span of the observation arc; Through the normal equation, the accuracy information of the four parameters of right ascension, declination, distance and distance change rate are calculated as follows: in, are the accuracy information of distance, distance change rate, right ascension and declination, σ ρ is the radar ranging accuracy, σ is the radar angle measurement accuracy, T is the time span of the observation arc, and N is the number of observation points.
8. The method according to claim 1, characterized in that In step 6, the allowable domain uncertainty error of the accuracy information is calculated by the Jacobian matrix: in, is the allowable uncertainty error of the accuracy information of right ascension, The theoretical radar observation data for the accuracy information of right ascension, The real radar observation data of the accuracy information of right ascension; is the Jacobian matrix of right ascension propagation; is the real arc attribute; is the actual arc attribute; is the allowable domain uncertainty error of the distance accuracy information, The theoretical radar observation data for distance accuracy information, Real radar observation data for distance accuracy information; is the Jacobian matrix of distance propagation; is the allowable uncertainty error of the accuracy information of declination, Theoretical radar observation data for the accuracy information of declination, Real radar observation data for the accuracy information of declination; is the Jacobian matrix of declination propagation; in, is the propagation function of right ascension; is the propagation function of declination; is the propagation function of distance; B is the arc attribute; is the actual arc attribute; ρ i is the distance at the i-th observation point, α i is the declination of the i-th observation point at the moment, δ i is the declination of the i-th observation point at the moment, x i is the orbital state at the i-th observation point; x0 is the orbital state at the initial moment.
9. The method according to claim 1, characterized in that In step 6, the minimum allowable domain error of the actual radar observation data is: in, is the minimum allowable domain error of the distance; is the actual radar observation distance of the ith object; is the i-th theoretical radar observation distance; represents normal distribution; is the distance observation error; is the Jacobian matrix of distance propagation; Σ B is the uncertainty matrix; is the minimum allowable domain error of right ascension; is the right ascension of the ith actual radar observation; is the right ascension of the ith theoretical radar observation; is the right ascension observation error; is the Jacobian matrix of right ascension propagation; is the minimum allowable domain error of declination; is the i-th actual radar observation declination; is the i-th theoretical radar observation declination; σ 2 is the angle observation error; is the Jacobian matrix of declination propagation; in, They are the accuracy information of distance, distance change rate, right ascension and declination respectively.
10. The method according to claim 1, characterized in that In step 7, the loss function constructed for: The loss function is used to calculate the loss function value of each sampling point in the original radar permissible domain, and the threshold β is set to ≥ 3; Eliminate the loss function value The sampling points that meet The sampling points of are included in the minimum radar allowable domain as the boundary of the minimum radar allowable domain, and the minimum radar allowable domain is constructed; Wherein, m is the total number of radar observation data; i is the sequence number of radar observation data; is the actual radar observation distance; is the theoretical radar observation distance; is the distance observation error; is the Jacobian matrix of the distance propagation corresponding to the least squares solution orbit; Σ B is the uncertainty matrix; is the actual radar observation right ascension; is the theoretical radar observation right ascension; is the right ascension observation error; is the Jacobian matrix of the right ascension propagation corresponding to the least squares solution orbit; is the actual radar observation declination; is the theoretical radar observation declination; σ 2 is the declination observation error; is the Jacobian matrix of the declination propagation corresponding to the least squares solution orbit.
