Low-orbit navigation enhanced hybrid constellation optimization method using non-dominated sorting genetic algorithm
Optimizing the dual Walker hybrid constellation through the non-dominant sorting genetic algorithm solves the problems of complex design and low-orbit satellite constellation design, achieving more efficient navigation coverage and accuracy, and improving the performance and reliability of GNSS.
Patent Information
- Application Number
- CN202411199074.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-29
- Publication Date
- 2024-12-06
AI Technical Summary
The existing low-orbit satellite constellation design has problems such as single goals, poor work efficiency and complex design, and it is difficult to optimize the constellation configuration and cannot meet the requirements of navigation enhancement systems for navigation accuracy and coverage.
The non-dominant sorting genetic algorithm is used to design a dual Walker hybrid constellation, and the constellation model is optimized through the multi-objective association chromosome non-dominant sorting genetic algorithm, optimize the orbital height and distribution of satellite constellations, and improve navigation performance and coverage performance.
It has achieved improved global navigation coverage performance for low-orbit satellite constellations, higher satellite utilization, higher navigation accuracy, and more efficient efficiency, enhancing the accuracy and reliability of GNSS.
Smart Images

Figure CN119089687A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of satellite navigation enhancement, and in particular relates to a low-orbit navigation enhancement hybrid constellation optimization method using a non-dominated sorting genetic algorithm. Background Art
[0002] With the upgrading of traditional industries and the development of new technologies such as 5G / 6G technology and the Internet of Things technology, users' requirements for navigation time and space accuracy and application scenarios are becoming more and more complex. A single Global Navigation Satellite System (GNSS) technology can no longer meet the service needs. Navigation augmentation technology is an important supplement to traditional GNSS, which can enhance the accuracy, stability, integrity, continuity and reliability of the GNSS system.
[0003] Existing navigation enhancement methods can be divided into two categories: satellite-based augmentation systems and ground-based augmentation systems. Among them, ground-based augmentation technology mainly builds reference stations on the ground to assist in correcting satellite signal errors, but it is limited by the distribution and coverage of ground equipment and cannot achieve global seamless coverage. Satellite-based augmentation technology is not limited by terrain and network transmission, and can achieve seamless high-precision positioning. It is the mainstream navigation enhancement idea currently used by various countries.
[0004] Low earth orbit (LEO) satellite constellations can support satellite communication network access and serve as navigation enhancement platforms for global navigation satellite systems due to their advantages such as low launch cost, short propagation delay, and low link loss. Before establishing a navigation enhancement platform, the design of satellite constellations is an important task. Previous LEO constellations usually consist of a single constellation configuration, which is not good for local navigation. At the same time, in the constellation design, the navigation enhancement task also has the problem of single consideration of goals, poor work efficiency, complex design, and difficulty in optimizing the constellation configuration. Therefore, it is necessary to propose a constellation layout method that can not only consider the constellation configuration of hybrid orbits, but also fully and comprehensively consider the performance of the enhanced system and can quickly optimize the design. Summary of the invention
[0005] The purpose of the present invention is to provide a low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, so that the navigation performance of GNSS is improved under the cooperation of a low-orbit enhanced satellite constellation, and at the same time, the navigation performance of the low-orbit satellite constellation is optimized under the constraints of satellite onboard equipment, space configuration and mission requirements, so that the constellation coverage performance is optimized and the navigation accuracy is higher.
[0006] The technical solution adopted by the present invention is to use a low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, which is specifically implemented in the following steps:
[0007] Step 1: Use the dual Walker hybrid constellation space configuration for design, establish a hybrid constellation model in the inertial coordinate system, discretize time and space, and establish a global observation model;
[0008] Step 2: Analyze the objective function and constraints to determine the constellation optimization mathematical model;
[0009] Step 3: Use the multi-objective associated chromosome non-dominated sorting genetic algorithm to optimize the constellation optimization mathematical model and obtain the global optimal low-orbit satellite constellation configuration.
[0010] The present invention is also characterized in that:
[0011] Step 1 is implemented as follows:
[0012] Step 1.1: The geometry of the Walker constellation is expressed in N w / P / F three integers, where N w is the total number of satellites in the constellation, P is the number of orbital planes, F is the phase factor, the first satellite on the first orbital plane in the constellation is defined as the seed satellite, and any satellite in the Walker constellation, that is, the jth satellite on the ith orbital plane, is expressed as: a ij =a0, e ij =e0、I ij =I0, ω ij =ω0,
[0013] The jth satellite in orbital plane i of the constellation is converted to the geocentric coordinate system, corresponding to the coordinates (x e ,y e ,z e ) is expressed as:
[0014]
[0015] Where: r e is the radius of the earth, R1, R2, R3 are coordinate transformation matrices;
[0016] Step 1.2: Discretize observation time and space and establish a global observation model:
[0017] Use the Fibonacci grid to generate uniform sampling points on the earth's surface and discretize the earth's observation surface. Suppose the number of sampling points is N. Then, in the geocentric coordinate system, the coordinates of the nth sampling point (x n ,y n ,z n ) is expressed as:
[0018]
[0019] Where: is the golden ratio;
[0020] For the orbital height a of the i-th Walker constellation i Under known conditions, the constellation period is known from Kepler's equation Where G is the gravitational constant of the Earth, M is the mass of the Earth, and the period T of the hybrid constellation is the least common multiple of T1 and T2, so T = lcm(T1, T2). The fixed discrete time sampling interval is Δt. For dual Walker constellations with different parameters, the number of sampling moments N is t =T / Δt;
[0021] Step 1.3: Construct decision variables based on the constellation model:
[0022] Assuming that the orbital heights of the dual Walker constellations are known, given the total number of satellites N total Under the premise of , the performance index of the constellation is affected by allocating the number of satellites of Walker constellations 1 and 2 and changing the orbital elements of their respective seed satellites. Let the decision variable x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2], where N i / P i / F i represents the geometry of the ith Walker constellation, N2 = N total -N1,I i and ω i Represents the orbital element of the i-th Walker constellation seed satellite.
[0023] Step 2 is implemented as follows:
[0024] Step 2.1: Analyze coverage performance and navigation performance and determine the objective function:
[0025] Objective function 1: global coverage:
[0026] Assume that the number of satellites visible above the nth sampling point at sampling time t is The minimum number of visible satellites for a navigation mission is v min , v min ≥4, the coverage state matrix of the sampling point at sampling time t The definition is as follows:
[0027]
[0028] Then the coverage C of the hybrid low-orbit navigation enhancement constellation in period T is expressed as:
[0029]
[0030] Objective function 2: Navigation accuracy:
[0031] Assume that at a certain moment, the ground receiver selects four satellites S1, S2, S3, and S4 for positioning solution. The position vector of the receiver in the geocentric inertial coordinate system is R o (X o ,Y o ,Z o ), the position vector from the satellite to the receiver is R i (X i ,Y i ,Z i ), where i = 1, 2, 3, 4, and the modulus length of the position vector from the satellite to the receiver is R i ;
[0032] Without considering the measurement noise, the position matrix is obtained:
[0033]
[0034] The DOP matrix D is obtained by transforming the position matrix:
[0035] D -1 =[P T P] -1 (6)
[0036] Decompose D into the following form:
[0037] D=[D ij ] (7)
[0038] The geometric dilution of precision (GDOP) is used as an indicator to measure the positioning accuracy of the navigation constellation. The GDOP solution is as follows:
[0039]
[0040] Assume that the GDOP observation matrix of observation point n at sampling time t is The definition is as follows:
[0041]
[0042] Where: is the GDOP observation value at observation point n at sampling time t;
[0043] The navigation performance G of the hybrid constellation of low-orbit satellites is expressed as:
[0044]
[0045] Step 2.2: Analyze the constraints:
[0046] Satellite elevation mask angle constraint: The satellite elevation mask angle ε, assuming that the minimum elevation mask angle of the satellite onboard equipment is ε min , then there is an elevation mask angle constraint:
[0047] ε≤ε min (11)
[0048] Continuous coverage constraint: Let the minimum continuous coverage time be T min , then the continuous coverage constraint is:
[0049]
[0050] Where: At any time t i ∈[t0,t last ] are satisfied
[0051] Precision failure constraint: D G When the value is greater than 10, the constellation positioning accuracy is poor and there are navigation accuracy constraints:
[0052]
[0053] Spatial configuration constraints: For the phase factor in the Walker constellation, there are the following constraints:
[0054] 0<F≤P-1 (14)
[0055] During the change process, N should always be maintained i With P i There is an integer multiple relationship between them, namely:
[0056] N i / P i =λ (15)
[0057] Where: λ∈N * ;
[0058] Step 2.3: Establish a hybrid low-orbit satellite navigation enhancement constellation optimization model:
[0059] 1) Maximize coverage
[0060]
[0061] 2) Minimize navigation precision factor
[0062]
[0063] In addition, the optimization process is subject to the following constraints:
[0064]
[0065] Step 3 is implemented as follows:
[0066] Step 3.1, Chaos initialization based on Cubic mapping:
[0067] The decision variables are binary coded, the chromosome length is L, the population size is C, and the actual value of each chromosome gene segment is r i j Indicates, where i=1…C represents the chromosome number in the population, and j=1…9 represents the position of the gene fragment in the chromosome;
[0068] The random generation rule of the population satisfies formula (19):
[0069]
[0070] in, represents the decision variable x j The upper and lower limits of the range; is a chaos operator, then the chaos calculation formula is expressed as:
[0071]
[0072] Among them, ρ is the control parameter. When ρ = 2.595, the generated chaotic variables have better ergodicity;
[0073] Step 3.2, calculate the non-dominated sorting and crowding degree of the initial population;
[0074] Step 3.3, Binary Tournament Selection;
[0075] Step 3.4, crossover of variant chromosomes. When the N1 / P1 / F1 variant chromosomes cross, the crossover process is first performed in chromosome N1. The N1 chromosome after the crossover is associated, and its corresponding P1 and F1 are expanded to form an intermediate chromosome. Then, on this basis, P1 and F1 are crossed according to the crossover probability to finally obtain a new chromosome.
[0076] Step 3.5: Adaptively calculate the mutation probability p according to the fitness function k , mutates according to probability. In genetic algorithms, p k The calculation formula should conform to the dynamic characteristics that change with the number of iterations k, and the expression is as follows:
[0077]
[0078] Step 3.6, using the mutation probability calculated by formula (21), for the mutation of the associated chromosome, Figure 2-(b) is the mutation strategy of the associated chromosome. When the N1 / P1 / F1 associated chromosome mutates, the mutated N1 will expand its associated chromosome, and the mutation of P1 will be carried out in the expanded associated chromosome, and the mutation of F1 is the same;
[0079] Step 3.7: Determine the termination condition and iterate.
[0080] The beneficial effects of the present invention are as follows: a hybrid constellation optimization method for low-orbit navigation enhancement using a non-dominated sorting genetic algorithm, and a hybrid constellation using a Walker configuration combination of low-orbit satellites, under the condition of a certain number of satellites, have better navigation coverage performance and higher satellite utilization rate on a global scale; under the premise of determining the configuration, the navigation coverage range and the navigation accuracy of the coverage interval of the navigation enhancement system are fully considered, and a multi-objective optimization model is constructed to enable the navigation enhancement system to have a wider coverage, more accurate navigation, and higher efficiency; a multi-objective associated chromosome non-dominated sorting genetic algorithm is formed by expanding the chromosome association in the genetic algorithm, initializing the population by using a Cubic chaotic map, and adaptively processing operators of the mutation process, and the algorithm can make the initial solution more random and ergodic, start the optimization process with a better initial value, reduce the number of constraint violations during the optimization process, adapt to the mutation law, and thereby improve the optimization efficiency and the quality of the approximate optimal solution; the method has obvious performance enhancement, can provide reliable and effective navigation backup information for GNSS, improves the accuracy, stability and reliability of GNSS, and provides a new reference and reference for the experimental development of the low-orbit satellite verification stage in my country, which is convenient for promotion and use. BRIEF DESCRIPTION OF THE DRAWINGS
[0081] Figure 1 It is a flowchart of the multi-objective associated chromosome non-dominated sorting genetic algorithm;
[0082] FIG2 is a schematic diagram of the crossover and mutation strategy of associated chromosomes in a multi-objective associated chromosome non-dominated sorting genetic algorithm;
[0083] FIG3 is a graph showing the approximate Pareto frontier surface results for different track height combinations in a specific embodiment;
[0084] FIG4 is a comparison diagram of the single-objective algorithm optimization process for different orbital height combinations in a specific embodiment;
[0085] FIG5 is a STK verification coverage diagram of approximate optimal solutions for different orbit height combinations in a specific embodiment. DETAILED DESCRIPTION
[0086] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.
[0087] The present invention uses a low-orbit navigation enhanced hybrid constellation optimization method of a non-dominated sorting genetic algorithm, analyzes the dual Walker hybrid low-orbit constellation spatial configuration to establish a constellation model and an observation model, uses the satellite visible matrix and the navigation accuracy matrix to give a multi-objective optimization function, and uses the satellite onboard equipment performance, coverage continuity, accuracy availability and spatial structure as constraints to construct a hybrid low-orbit satellite navigation enhanced constellation optimization model. According to the established multi-objective optimization model, a multi-objective associated chromosome non-dominated sorting genetic algorithm is designed to solve the problems of randomness of pseudo-random initial solutions, poor traversal, and strict spatial configuration constraints, thereby avoiding the waste of resources due to strong constraints and accelerating convergence efficiency, so that the constellation design can find a better spatial configuration at a lower cost and faster speed. It provides a reliable and valuable reference solution for the deployment of low-orbit navigation enhanced constellations.
[0088] The present invention uses a low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, and the flow chart is as follows: Figure 1 As shown, the specific implementation steps are as follows:
[0089] Step 1: Use the dual Walker hybrid constellation space configuration for design, propose decision variables x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2] for the configuration, establish a hybrid constellation model in the inertial coordinate system, discretize time and space, and establish a global observation model;
[0090] Step 1 is implemented as follows:
[0091] Step 1.1: Analyze the orbital configuration of the Walker constellation's low-orbit satellites, determine a dual-Walker hybrid constellation space configuration for constellation design, and establish a constellation model.
[0092] The geometry of the Walker constellation is shown in N w / P / F three integers, where N w is the total number of satellites in the constellation, P is the number of orbital planes, F is the phase factor, the first satellite on the first orbital plane in the constellation is defined as the seed satellite, and any satellite in the Walker constellation, that is, the jth satellite on the ith orbital plane, is expressed as: a ij =a0, e ij =e0、I ij =I0, ω ij =ω0,
[0093]
[0094] The jth satellite in orbital plane i of the constellation is converted to the geocentric coordinate system, corresponding to the coordinates (x e ,y e ,z e) is expressed as:
[0095]
[0096] Where: r e is the radius of the earth, R1, R2, R3 are coordinate transformation matrices;
[0097] Step 1.2: Discretize observation time and space and establish a global observation model:
[0098] Use the Fibonacci grid to generate uniform sampling points on the earth's surface and discretize the earth's observation surface. Suppose the number of sampling points is N. Then, in the geocentric coordinate system, the coordinates of the nth sampling point (x n ,y n ,z n ) is expressed as:
[0099]
[0100] Where: is the golden ratio;
[0101] For the orbital height a of the i-th Walker constellation i Under known conditions, the constellation period is known from Kepler's equation Where G is the gravitational constant of the Earth, M is the mass of the Earth, and the period T of the hybrid constellation is the least common multiple of T1 and T2, so T = lcm(T1, T2). The fixed discrete time sampling interval is Δt. For dual Walker constellations with different parameters, the number of sampling moments N is t =T / Δt;
[0102] Step 1.3: Construct decision variables based on the constellation model:
[0103] The configuration of the dual Walker constellation determines its coverage and navigation performance. Assuming that the orbital heights of the dual Walker constellation are known, given the total number of satellites N total Under the premise of , the performance index of the constellation is affected by allocating the number of satellites of Walker constellations 1 and 2 and changing the orbital elements of their respective seed satellites. Let the decision variable x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2], where N i / P i / F i represents the geometry of the ith Walker constellation, N2 = N total -N1,I i and ω i Represents the orbital element of the i-th Walker constellation seed satellite.
[0104] Step 2: Analyze the objective function and constraints, and determine the mathematical model for constellation optimization. Take the global coverage rate and the highest navigation accuracy of the coverage range of the hybrid constellation as the mission requirements, establish an optimization model, and the objective function is the average coverage rate C and the average navigation accuracy factor G. The constraints include satellite elevation mask angle constraints, space configuration constraints, etc.
[0105] Step 2 is implemented as follows:
[0106] Step 2.1: Analyze coverage performance and navigation performance and determine the objective function:
[0107] Objective function 1: global coverage:
[0108] For satellite navigation to work properly, the sky above the ground observation point must be covered by at least 4 satellites at the same time. Suppose the number of satellites visible above the nth sampling point at sampling time t is The minimum number of visible satellites for a navigation mission is v min , v min ≥4, the coverage state matrix of the sampling point at sampling time t The definition is as follows:
[0109]
[0110] Then the coverage C of the hybrid low-orbit navigation enhancement constellation in period T is expressed as:
[0111]
[0112] Objective function 2: Navigation accuracy:
[0113] Assume that at a certain moment, the ground receiver selects four satellites S1, S2, S3, and S4 for positioning solution. The position vector of the receiver in the geocentric inertial coordinate system is R o (X o ,Y o ,Z o ), the position vector from the satellite to the receiver is R i (X i ,Y i ,Z i ), where i = 1, 2, 3, 4, and the modulus length of the position vector from the satellite to the receiver is R i ;
[0114] Without considering the measurement noise, the position matrix is obtained:
[0115]
[0116] The DOP matrix D is obtained by transforming the position matrix:
[0117] D -1 =[PT P] -1 (6)
[0118] Decompose D into the following form:
[0119] D=[D ij ] (7)
[0120] The geometric dilution of precision (GDOP) is used as an indicator to measure the positioning accuracy of the navigation constellation. The GDOP solution is as follows:
[0121]
[0122] Assume that the GDOP observation matrix of observation point n at sampling time t is The definition is as follows:
[0123]
[0124] Where: is the GDOP observation value at observation point n at sampling time t;
[0125] The navigation performance G of the hybrid constellation of low-orbit satellites is expressed as:
[0126]
[0127] Step 2.2: Analyze the constraints:
[0128] Satellite elevation mask angle constraint: The satellite elevation mask angle ε determines the coverage of a single satellite on the ground. Suppose the minimum elevation mask angle of the satellite onboard equipment is ε min , then there is an elevation mask angle constraint:
[0129] ε≤ε min (11)
[0130] Continuous coverage constraint: Let the minimum continuous coverage time be T min , then the continuous coverage constraint is:
[0131]
[0132] Where: At any time t i ∈[t0,t last ] are satisfied
[0133] Precision failure constraint: D G When the value is greater than 10, the constellation positioning accuracy is poor and there are navigation accuracy constraints:
[0134]
[0135] Spatial configuration constraints: For the phase factor in the Walker constellation, there are the following constraints:
[0136] 0<F≤P-1 (14)
[0137] During the change process, N should always be maintained i With P i There is an integer multiple relationship between them, namely:
[0138] N i / P i =λ (15)
[0139] Where: λ∈N * ;
[0140] Step 2.3: Establish a hybrid low-orbit satellite navigation enhancement constellation optimization model:
[0141] 1) Maximize coverage
[0142]
[0143] 2) Minimize navigation precision factor
[0144]
[0145] In addition, the optimization process is subject to the following constraints:
[0146]
[0147] Step 3: Use the multi-objective associated chromosome non-dominated sorting genetic algorithm to optimize the constellation optimization mathematical model and obtain the global optimal low-orbit satellite constellation configuration. The process is as follows: Binary initial encoding of the spatial configuration decision variables is performed according to the Cubic chaotic mapping principle to obtain the gene encoding of the corresponding position; for the strong constraint relationship of the optimization problem, the chromosomes in the genetic algorithm are expanded in association, the associated chromosomes are constructed, and the crossover and mutation strategies of the associated chromosomes are designed; the fitness function is calculated, the population is selected, and the offspring genes are crossed; the mutation probability p is adaptively calculated according to the fitness function k , mutate according to probability; merge the population and sort the non-dominated population and calculate the population crowding degree.
[0148] Step 3 is implemented as follows:
[0149] Step 3.1, Chaos initialization based on Cubic mapping:
[0150] The decision variables are binary coded, the chromosome length is L, the population size is C, and the actual value of each chromosome gene segment is r i jIndicates, where i=1…C represents the chromosome number in the population, and j=1…9 represents the position of the gene fragment in the chromosome;
[0151] The random generation rule of the population satisfies formula (19):
[0152]
[0153] in, represents the decision variable x j The upper and lower limits of the range; is a chaos operator, then the chaos calculation formula is expressed as:
[0154]
[0155] Among them, ρ is the control parameter. When ρ = 2.595, the generated chaotic variables have better ergodicity;
[0156] Step 3.2, calculate the non-dominated sorting and crowding degree of the initial population;
[0157] Step 3.3, Binary Tournament Selection;
[0158] Step 3.4, crossover of variant chromosomes. Figure 2-(a) shows the crossover strategy of associated chromosomes. When the N1 / P1 / F1 variant chromosomes cross, the crossover process is first performed in chromosome N1. The N1 chromosome after the crossover is associated, and its corresponding P1 and F1 are expanded to form an intermediate chromosome. Then, on this basis, P1 and F1 are crossed according to the crossover probability to finally obtain a new chromosome.
[0159] Step 3.5: Adaptively calculate the mutation probability p according to the fitness function k , mutates according to probability. In genetic algorithms, p k The calculation formula should conform to the dynamic characteristics that change with the number of iterations k, and the expression is as follows:
[0160]
[0161] Step 3.6, using the mutation probability calculated by formula (21), for the mutation of the associated chromosome, Figure 2-(b) is the mutation strategy of the associated chromosome. When the N1 / P1 / F1 associated chromosome mutates, the mutated N1 will expand its associated chromosome, and the mutation of P1 will be carried out in the expanded associated chromosome, and the mutation of F1 is the same;
[0162] Step 3.7: Determine the termination condition and iterate.
[0163] Example 1
[0164] The present invention uses a low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, and the flow chart is as follows: Figure 1 As shown, the specific implementation steps are as follows:
[0165] Step 1: Use the dual Walker hybrid constellation space configuration for design, propose decision variables x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2] for the configuration, establish a hybrid constellation model in the inertial coordinate system, discretize time and space, and establish a global observation model;
[0166] Step 2: Analyze the objective function and constraints, and determine the mathematical model for constellation optimization. Take the global coverage rate and the highest navigation accuracy of the coverage range of the hybrid constellation as the mission requirements, establish an optimization model, and the objective function is the average coverage rate C and the average navigation accuracy factor G. The constraints include satellite elevation mask angle constraints, space configuration constraints, etc.
[0167] Step 3: Use the multi-objective associated chromosome non-dominated sorting genetic algorithm to optimize the constellation optimization mathematical model and obtain the global optimal low-orbit satellite constellation configuration. The process is as follows: Binary initial encoding of the spatial configuration decision variables is performed according to the Cubic chaotic mapping principle to obtain the gene encoding of the corresponding position; for the strong constraint relationship of the optimization problem, the chromosomes in the genetic algorithm are expanded in association, the associated chromosomes are constructed, and the crossover and mutation strategies of the associated chromosomes are designed; the fitness function is calculated, the population is selected, and the offspring genes are crossed; the mutation probability p is adaptively calculated according to the fitness function k , mutate according to probability; merge the population and sort the non-dominated population and calculate the population crowding degree.
[0168] Example 2
[0169] The present invention uses a low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, and the flow chart is as follows: Figure 1 As shown, the specific implementation steps are as follows:
[0170] Step 1: Use the dual Walker hybrid constellation space configuration for design, propose decision variables x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2] for the configuration, establish a hybrid constellation model in the inertial coordinate system, discretize time and space, and establish a global observation model;
[0171] Step 1 is implemented as follows:
[0172] Step 1.1: Analyze the orbital configuration of the Walker constellation's low-orbit satellites, determine a dual-Walker hybrid constellation space configuration for constellation design, and establish a constellation model.
[0173] The geometry of the Walker constellation is represented by N w / P / F three integers, where N wis the total number of satellites in the constellation, P is the number of orbital planes, F is the phase factor, the first satellite on the first orbital plane in the constellation is defined as the seed satellite, and any satellite in the Walker constellation, that is, the jth satellite on the ith orbital plane, is expressed as: a ij =a0, e ij =e0、I ij =I0, ω ij =ω0,
[0174]
[0175] The jth satellite in orbital plane i of the constellation is converted to the geocentric coordinate system, corresponding to the coordinates (x e ,y e ,z e ) is expressed as:
[0176]
[0177] Where: r e is the radius of the earth, R1, R2, R3 are coordinate transformation matrices;
[0178] Step 1.2: Discretize observation time and space and establish a global observation model:
[0179] Use the Fibonacci grid to generate uniform sampling points on the earth's surface and discretize the earth's observation surface. Suppose the number of sampling points is N. Then, in the geocentric coordinate system, the coordinates of the nth sampling point (x n ,y n ,z n ) is expressed as:
[0180]
[0181] Where: is the golden ratio;
[0182] For the orbital height a of the i-th Walker constellation i Under known conditions, the constellation period is known from Kepler's equation Where G is the gravitational constant of the Earth, M is the mass of the Earth, and the period T of the hybrid constellation is the least common multiple of T1 and T2, so T = lcm(T1, T2). The fixed discrete time sampling interval is Δt. For dual Walker constellations with different parameters, the number of sampling moments N is t =T / Δt;
[0183] Step 1.3: Construct decision variables based on the constellation model:
[0184] The configuration of the dual Walker constellation determines its coverage and navigation performance. Assuming that the orbital heights of the dual Walker constellation are known, given the total number of satellites N total Under the premise of , the performance index of the constellation is affected by allocating the number of satellites of Walker constellations 1 and 2 and changing the orbital elements of their respective seed satellites. Let the decision variable x = [N1, P1, F1, I1, ω1, P2, F2, I2, ω2], where N i / P i / F i represents the geometry of the ith Walker constellation, N2 = N total -N1,I i and ω i Represents the orbital element of the i-th Walker constellation seed satellite.
[0185] Step 2: Analyze the objective function and constraints to determine the constellation optimization mathematical model;
[0186] Step 3: Use the multi-objective associated chromosome non-dominated sorting genetic algorithm to optimize the constellation optimization mathematical model and obtain the global optimal low-orbit satellite constellation configuration.
[0187] Step 3 is implemented as follows:
[0188] Step 3.1, Chaos initialization based on Cubic mapping:
[0189] The decision variables are binary coded, the chromosome length is L, the population size is C, and the actual value of each chromosome gene segment is r i j Indicates, where i=1…C represents the chromosome number in the population, and j=1…9 represents the position of the gene fragment in the chromosome;
[0190] The random generation rule of the population satisfies formula (19):
[0191]
[0192] in, represents the decision variable x j The upper and lower limits of the range; is a chaos operator, then the chaos calculation formula is expressed as:
[0193]
[0194] Among them, ρ is the control parameter. When ρ = 2.595, the generated chaotic variables have better ergodicity;
[0195] Step 3.2, calculate the non-dominated sorting and crowding degree of the initial population;
[0196] Step 3.3, Binary Tournament Selection;
[0197] Step 3.4, crossover of variant chromosomes. Figure 2-(a) shows the crossover strategy of associated chromosomes. When the N1 / P1 / F1 variant chromosomes cross, the crossover process is first performed in chromosome N1. The N1 chromosome after the crossover is associated, and its corresponding P1 and F1 are expanded to form an intermediate chromosome. Then, on this basis, P1 and F1 are crossed according to the crossover probability to finally obtain a new chromosome.
[0198] Step 3.5: Adaptively calculate the mutation probability p according to the fitness function k , mutates according to probability. In genetic algorithms, p k The calculation formula should conform to the dynamic characteristics that change with the number of iterations k, and the expression is as follows:
[0199]
[0200] Step 3.6, using the mutation probability calculated by formula (21), for the mutation of the associated chromosome, Figure 2-(b) is the mutation strategy of the associated chromosome. When the N1 / P1 / F1 associated chromosome mutates, the mutated N1 will expand its associated chromosome, and the mutation of P1 will be carried out in the expanded associated chromosome, and the mutation of F1 is the same;
[0201] Step 3.7: Determine the termination condition and iterate.
[0202] Example 3
[0203] The dual Walker hybrid orbit is used to optimize the design of the low-orbit satellite navigation enhancement system. The total number of satellites is N total There are 80 satellites, and the orbital altitude combinations are divided into four groups: 600+800km, 700+900km, 800+1000km, and 900+1100km. The ranges of the decision variables are as follows: the total number of Walker1 constellation satellites N1 is 20 to 60, the number of orbital planes P1 is 2 to 9, the phase factor F1 is 1 to P1-1, the orbital inclination of the seed satellite I1 is 0 to 90°, and the mean anomaly ω1 is 0 to 2πrad, the number of orbital planes of the Walker2 constellation P2 is 2 to 9, the phase factor F2 is 1 to P2-1, the orbital inclination of the seed satellite I2 is 0 to 90°, and the mean anomaly ω2 is 0 to 2πrad. Adopt Figure 1 The algorithm in the flowchart solves the multi-objective optimization problem with a population size of 15, 120 iterations, a crossover probability of 0.2, a mutation factor of 0.6, and a Pareto weight of 10. The observation model mission time is the hybrid orbit period, the number of spatial sampling points is 200, and the time sampling interval is 1 minute. The approximate Pareto frontier of different orbit combinations is as follows Figure 3(a) to Figure 3(d)As shown in Figure 2, compared with the traditional single-objective genetic algorithm (GA) and particle swarm algorithm (PSO) optimization process, Figure 4(a) to Figure 4(d) After optimization, STK verification covers Figure 5(a) to Figure 5(d) .
[0204] [1]LI B,GE H,GE M,et al.LEO enhanced Global Navigation SatelliteSystem(LeGNSS) for real-time precise positioning services[J].Advances in spaceresearch,2019,63(1):73-93.
[0205] [2]DEL PORTILLO I,CAMERON BG,CRAWLEY E FA technical comparison of three low earth orbit satellite constellation systems to provide globalbroadband[J]. Acta astronautica, 2019,159:123-135.
[0206] [3]ASGARIMEHR M,HOSSAINALI M M.Optimization of geosynchronous satellite constellation for independent regional navigation and positioningin Middle East region[J].Acta Astronautica, 2014,104(1):147-158.
[0207] [4]ARIAS-MONTANO A,COELLO CAC,MEZURA-MONTES E.Multi-objectiveevolutionary algorithms in aeronautical and aerospace engineering[J].IEEE transactions on evolutionary computation, 2012,16(5):662-694.
[0208] [5]YIJIE S,GONGZHANG S.Improved NSGA-II multi-objective genetic algorithm based on hybridization-encouraged mechanism[J].Chinese Journal ofAeronautics,2008,21(6):540-549.
[0209] [6]YI HAN,LEI W,WENJU FU,et al.LEO navigation augmentationconstellation design with the multi-objective optimization approaches[J].Chinese Journal of Aeronautics,2021,34(4):265-278.
[0210] [7] Li Shuang, Gong Yifei, Li Kexing, et al. Distributed autonomous relative navigation method for time-varying topology satellite clusters[J]. Journal of Astronautics, 2023, 44(01): 119-131.
[0211] [8]LIU A,XU
[0212] [9] LITTORIANO G, COLOMBO C, NASTASI A, et al. Modeling of spacecraftapparent brightness A study on OneWeb constellation satellites[J]. Advances in Space Research, 2024.
[0213]
[10] GUAN M,
Claims
1. Low-orbit navigation enhanced hybrid constellation optimization method using non-dominated sorting genetic algorithm, It is characterized in that Follow the steps below to implement it: Step 1: Use the dual Walker hybrid constellation space configuration for design, establish a hybrid constellation model in the inertial coordinate system, discretize time and space, and establish a global observation model; Step 2: Analyze the objective function and constraints to determine the constellation optimization mathematical model; Step 3: Use the multi-objective associated chromosome non-dominated sorting genetic algorithm to optimize the constellation optimization mathematical model and obtain the global optimal low-orbit satellite constellation configuration.
2. According to claim 1, the low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm, It is characterized in that The step 1 is specifically implemented according to the following steps: Step 1.1: The geometry of the Walker constellation is expressed in N w / P / F three integers, where N w is the total number of satellites in the constellation, P is the number of orbital planes, F is the phase factor, the first satellite on the first orbital plane in the constellation is defined as the seed satellite, and any satellite in the Walker constellation, that is, the jth satellite on the ith orbital plane, is expressed as: a ij =a 0 、e ij =e 0 ,I ij =I 0 , ω ij =ω 0 , The jth satellite in orbital plane i of the constellation is converted to the geocentric coordinate system, corresponding to the coordinates (x e ,y e ,z e ) is expressed as: Where: r e is the radius of the Earth, R 1 , R 2 , R 3 is the coordinate transformation matrix; Step 1.2: Discretize observation time and space and establish a global observation model: Use the Fibonacci grid to generate uniform sampling points on the earth's spherical surface and discretize the earth's observation surface. Suppose the number of sampling points is N. Then in the geocentric coordinate system, the coordinates of the nth sampling point (x n ,y n ,z n ) is expressed as: Where: is the golden ratio; For the orbital height a of the i-th Walker constellation i Under known conditions, the constellation period is known from Kepler's equation Where G is the Earth’s gravitational constant, M is the Earth’s mass, and the hybrid constellation period T is T 1 With T 2 The lowest common multiple of T = lcm(T 1 ,T 2 ), the discrete time sampling interval is fixed to Δt, then for the dual Walker constellation with different parameters, the number of sampling moments N t =T / Δt; Step 1.3: Construct decision variables based on the constellation model: Assuming that the orbital heights of the dual Walker constellations are known, given the total number of satellites N total Under the premise of , by allocating the number of satellites of Walker constellation 1 and 2 and changing the orbital elements of each seed satellite, the performance index of the constellation is affected. Let the decision variable x = [N 1 ,P 1 ,F 1 ,I 1 ,ω 1 ,P 2 ,F 2 ,I 2 ,ω 2 ], where N i / P i / F i represents the geometry of the ith Walker constellation, N 2 =N total -N 1 , I i and ω i Represents the orbital element of the i-th Walker constellation seed satellite.
3. The low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm according to claim 2, It is characterized in that The step 2 is specifically implemented according to the following steps: Step 2.1, analyze coverage performance and navigation performance, and determine the objective function; Step 2.2: Analyze the constraints: Satellite elevation mask angle constraint: The satellite elevation mask angle ε, assuming that the minimum elevation mask angle of the satellite onboard equipment is ε min , then there is an elevation mask angle constraint: ε≤ε min (11) Continuous coverage constraint: Let the minimum continuous coverage time be T min , then the continuous coverage constraint is: Where: At any time t i ∈[t 0 ,t last ] are satisfied Precision failure constraint: D G When the value is greater than 10, the constellation positioning accuracy is poor and there are navigation accuracy constraints: Spatial configuration constraints: For the phase factor in the Walker constellation, there are the following constraints: 0<F≤P-1 (14)(1)N should always be maintained during the change process i With P i There is an integer multiple relationship between them, namely: N i / P i =λ (15) Where: λ∈N * ; Step 2.3: Establish a hybrid low-orbit satellite navigation enhancement constellation optimization model: 1) Maximize coverage (2) Minimize the navigation precision factor In addition, the optimization process is subject to the following constraints:
4. The low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm according to claim 3, It is characterized in that The step 2.1 is as follows: Objective function 1: global coverage: Assume that the number of satellites visible above the nth sampling point at sampling time t is The minimum number of visible satellites for a navigation mission is v min , v min ≥4, the coverage state matrix of the sampling point at sampling time t The definition is as follows: Then the coverage C of the hybrid low-orbit navigation enhancement constellation in period T is expressed as: Objective function 2: Navigation accuracy: Assume that at a certain moment, the ground receiver selects four satellites S for positioning solution 1 ,S 2 ,S 3 ,S 4 The position vector of the receiver in the geocentric inertial coordinate system is R o (X o ,Y o ,Z o ), the position vector from the satellite to the receiver is R i (X i ,Y i ,Z i ), where i = 1, 2, 3, 4, and the modulus length of the position vector from the satellite to the receiver is R i ; Without considering the measurement noise, the position matrix is obtained: The DOP matrix D is obtained by transforming the position matrix: D -1 =[P T P] -1 (6) Decompose D into the following form: D=[D ij ] (7) The geometric dilution of precision (GDOP) is used as an indicator to measure the positioning accuracy of the navigation constellation. The GDOP solution is as follows: Assume that the GDOP observation matrix of observation point n at sampling time t is The definition is as follows: Where: is the GDOP observation value at observation point n at sampling time t; The navigation performance G of the hybrid constellation of low-orbit satellites is expressed as:
5. The low-orbit navigation enhanced hybrid constellation optimization method using a non-dominated sorting genetic algorithm according to claim 4, It is characterized in that The step 3 is specifically implemented according to the following steps: Step 3.1, Chaos initialization based on Cubic mapping: The decision variables are binary coded, the chromosome length is L, the population size is C, and the actual value of each chromosome gene segment is r i j Indicates, where i=1…C represents the chromosome number in the population, and j=1…9 represents the position of the gene fragment in the chromosome; The population random generation rule satisfies formula (19): in, represents the decision variable x j The upper and lower limits of the range; is a chaos operator, then the chaos calculation formula is expressed as: Among them, ρ is the control parameter. When ρ = 2.595, the generated chaotic variables have better ergodicity; Step 3.2, calculate the non-dominated sorting and crowding degree of the initial population; Step 3.3, Binary Tournament Selection; Step 3.4: Crossover of mutant chromosomes, when N 1 / P 1 / F 1 When the mutant chromosome crosses over, the crossover process first occurs on chromosome N. 1 After the crossover, N 1 Chromosomes are associated and their corresponding P 1 With F 1 , forming an intermediate chromosome, and then on this basis 1 With F 1 Crossover is performed according to the crossover probability, and finally a new chromosome is obtained; Step 3.5: Adaptively calculate the mutation probability p according to the fitness function k , mutates according to probability. In genetic algorithms, p k The calculation formula should conform to the dynamic characteristics that change with the number of iterations k, and the expression is as follows: Step 3.6, the mutation probability calculated by formula (21) is used to mutate the associated chromosome. Figure 2-(b) shows the mutation strategy of the associated chromosome. When N 1 / P 1 / F 1 When the chromosome mutation is associated, the N 1 will expand its associated chromosome, P 1 The mutation will be carried out in the expanded associated chromosome, F 1 The same applies to the variation of Step 3.7: Determine the termination condition and iterate.
