A high-fidelity spacecraft high-precision atmospheric drag modeling method

Abstract

This invention belongs to the field of spacecraft modeling methods and discloses a high-fidelity, high-precision atmospheric drag modeling method for spacecraft. The method includes steps such as constructing a 3D spacecraft model database, initializing the computational domain and setting state update rules, simulating the interaction between test particles and the spacecraft surface, and user algorithms for the model. This invention utilizes the Monte Carlo simulation method to simulate the interaction between test particles and the spacecraft surface, and is applicable to spacecraft of arbitrary shapes in sparse atmospheric environments, demonstrating strong applicability. In the simulation, test particles are added to the computational domain one by one, minimizing computation time and storage space consumption. A grid file is used to store the atmospheric drag modeling results, facilitating users to obtain atmospheric drag acceleration at any position and attitude. This invention has broad application prospects in precise orbit determination of low Earth orbit spacecraft and high-precision satellite positioning.

Description

Technical Field

[0001] This invention belongs to the field of spacecraft modeling methods, specifically relating to a high-fidelity, high-precision atmospheric drag modeling method for spacecraft. Background Technology

[0002] With increasing demands for high-precision positioning of spacecraft within specific coordinate frames in space missions, accurate spacecraft orbital mechanics models have become an engineering necessity. Taking Low Earth Orbit (LEO) satellites as an example, a fundamental prerequisite for providing communication, navigation, and remote sensing services is achieving real-time precise orbit determination and time synchronization. The orbits of LEO satellites are affected by non-conservative forces such as atmospheric drag and radiation pressure. These non-conservative forces are difficult to model and predict accurately, posing a challenge to real-time precise orbit determination for LEO satellites.

[0003] Compared to the relatively mature and accurate modeling techniques for Earth's gravity and lunar gravity, non-conservative forces such as atmospheric drag and radiation pressure are smaller in magnitude, but they can significantly impact the orbits of long-term spacecraft, especially low-Earth orbit (LEO) satellites located in the free molecular flow region of the atmosphere. Atmospheric drag acts in the opposite direction to the relative motion between the satellite and the atmosphere, primarily reducing the satellite's velocity. Due to the reduction in mechanical energy, the satellite's altitude gradually decreases over time under the influence of atmospheric drag. For example, due to miscalculations in the design and manufacturing of Starlink satellites, the propulsion system was unable to compensate for the effects of atmospheric drag during geomagnetic storms, leading to the destruction of numerous satellites and irreparable economic losses. Therefore, accurate modeling and precise orbit determination of atmospheric drag for LEO satellites in complex space environments are among the key technologies that urgently need to be addressed in the satellite manufacturing and mission scheduling processes of large-scale LEO constellations.

[0004] According to the free molecular flow theory, atmospheric molecules undergo mirror reflection, diffuse scattering, and retention before and after colliding with the surface of a low Earth orbit spacecraft. The general model of atmospheric drag is as follows:

[0005]

[0006] Where F represents atmospheric drag; Cd is the drag coefficient; A is the effective windward area; ρ is the atmospheric density; and Vr is the relative velocity. From the above equation, it can be seen that the main uncertainties in the atmospheric drag model come from atmospheric density, the satellite's effective windward area, and the drag coefficient. That is, the accurate modeling of atmospheric drag is mainly related to the accuracy of the atmospheric density model and the atmosphere-satellite interaction model.

[0007] Currently, widely used atmospheric density models include the American Standard Atmospheric Model, the Jacchia series of atmospheric models, the MSIS series of models, the JB-2008 model, and the DTM series of models. The interaction between atmospheric molecules and the satellite surface is complex. Depending on the temperature and kinetic energy of the atmospheric molecules, they can adhere to the satellite surface or be bounced off. Currently, in atmospheric drag modeling, there are two main methods for handling the effective windward area of ​​the satellite. One method is to select one-quarter of the satellite's surface area and treat it as the effective windward area; the other is to treat the effective windward area as a parameter and estimate it along with the orbital state during satellite operation. The atmospheric drag coefficient is related to many factors, including atmospheric composition, satellite surface material, and shape. Accurately determining the atmospheric drag coefficient is very difficult. Generally, the drag coefficient is treated as a parameter and estimated along with the orbital state, which can, to some extent, correct the errors caused by atmospheric density in the atmospheric drag model, or it can be solved using drag coefficient formulas (such as the COOK formula).

[0008] Besides the analytical methods used to calculate atmospheric drag, Monte Carlo simulation can also be used. Direct Simulation Monte Carlo (DSMC) simulates dilute gas flow at the individual molecular level. The airflow is represented by a set of simulated particles, each considered a representative of real molecules in the airflow. Any dynamic processes in the gas are considered as changes in state variables of individual particles due to collisions, free motion, and interactions with boundaries, thus using physical simulation to calculate the atmospheric drag experienced by the satellite. Test Particle Monte Carlo (TPMC) is characterized by the sequential rather than simultaneous generation of simulated molecules; that is, only one test particle is generated at a time. Therefore, it does not consume too much computation time and storage space, making it more suitable for flow problems involving multiple surface reflections due to complex boundaries.

[0009] Furthermore, the physical mechanisms of atmospheric drag indicate that atmospheric drag models are specific to different spacecraft, making it impossible to create a model applicable to all spacecraft. Different models are needed for different spacecraft, increasing the complexity of the modeling process. Therefore, it is essential to research a universal physical modeling method for spacecraft atmospheric drag that is highly applicable and accurate. The simulation process also needs to be optimized to achieve faster computational efficiency. Summary of the Invention

[0010] To address the aforementioned technical problems, this invention provides a high-fidelity, high-precision atmospheric drag modeling method for spacecraft, thereby resolving the issues in the prior art. To achieve the aforementioned objective, the technical solution adopted by this invention is as follows:

[0011] A high-fidelity, high-precision atmospheric drag modeling method for spacecraft includes the following steps:

[0012] The first step is to construct a 3D model database for the spacecraft: the basic geometric prototype of the spacecraft is described in a parametric form, and a 3D model of the spacecraft is constructed by combining the basic geometric prototype; the 3D model database stores the geometric parameters and surface material properties of each component of the 3D model of the spacecraft.

[0013] The second step is to initialize the computational domain and set the state update rules: construct a computational domain centered on the three-dimensional model of the spacecraft; simulate the test particle flow within the computational domain and set the initial state of the test particle flow;

[0014] The third step is to simulate and test the interaction between the particles and the spacecraft surface: using the Monte Carlo simulation method, the interaction between the particles and the spacecraft surface is simulated and tested; after multiple iterations and stabilization, the product of the atmospheric drag coefficient in each direction and the spacecraft surface area is output.

[0015] The fourth step is the user algorithm for the model: After completing the atmospheric drag calculation in all directions, the product of the atmospheric drag coefficient and the surface area of ​​the spacecraft at any position is output in the spacecraft coordinate system. This allows the atmospheric drag acceleration of the three-dimensional model of the spacecraft at any position and in any attitude to be calculated.

[0016] Furthermore, in the first step, geometric models of each component of the spacecraft are performed based on the actual shape of the satellite, and the constructed three-dimensional model database of the spacecraft is reorganized using a KD-tree data structure.

[0017] Furthermore, the second step also includes setting state update rules, which include at least three states of the test particles: no collision and still within the computational domain, no collision but flying out of the computational domain, and colliding with the spacecraft surface.

[0018] Furthermore, in the second step, setting the initial state of the test particle flow includes the following steps:

[0019] Simulate the test particle flow within the computational domain and set the initial position X0 and initial velocity V0 of the test particle flow;

[0020] The test particle stream consists of several test particles, the number of which follows an atmospheric density model and the distribution follows a Boltzmann distribution. The initial position X0 is calculated. The initial velocity of the test particle stream includes the thermal velocity V.m and relative velocity V r ;

[0021] If the test particle flow obeys thermal motion, the atmospheric inflow temperature T can be obtained through an atmospheric density model. ∞ The thermal velocity of the tested particle is:

[0022]

[0023] In the formula, m is the mass of a single test particle;

[0024] If the spacecraft's position remains unchanged, the test particle moves at a relative velocity V. r sports;

[0025] Set the time iteration step size Δt, and assume that the test particles move in a straight line within the iteration step size. Update the positions X of all test particles. 0+1 and speed V 0+1 And determine whether a collision or flight out of the computational domain has occurred;

[0026] If the test particle collides with the spacecraft surface, update the velocity of the test particle after the collision, and record the rate of change of momentum Δp at this time.

[0027] If the test particle flies out of the computational domain without colliding with anything, then a new test particle is placed in the computational domain.

[0028] If the test particle does not collide with anything and does not fly out of the computational domain, then the position and velocity of the test particle remain P. 0+1 and V 0+1 .

[0029] Furthermore, in the third step, when simulating the interaction between the particles and the spacecraft surface using the Monte Carlo simulation method, the following steps are included:

[0030] The Monte Carlo simulation method was used to simulate the interaction between particles and the spacecraft surface, and the rate of change of momentum Δp at the time of collision was recorded. i ;

[0031] After multiple iterations and stabilization, the atmospheric drag in this direction was obtained. Then we can obtain the product C of the atmospheric drag coefficient and the spacecraft surface area. d A;

[0032] Set the grid resolution to get the total number of directions, and adjust the spacecraft's position in the computational domain;

[0033] Repeat the above steps to perform Monte Carlo simulations in other directions and output the product of atmospheric drag coefficients and spacecraft surface area in those other directions.

[0034] Furthermore, in the fourth step, the product of the atmospheric drag coefficient and the spacecraft surface area at any location is output in the form of a grid file. After obtaining the grid file, the user can use interpolation to calculate the atmospheric drag acceleration at any location and in any attitude.

[0035] The present invention has the following beneficial effects:

[0036] (1) This invention simulates the interaction between test particles and the spacecraft surface using the Monte Carlo simulation method for test particles. It is applicable to spacecraft of arbitrary shapes in sparse atmospheric environments, has strong applicability, and can be flexibly used for modeling different spacecraft. As a non-conservative force modeling method, it can improve and enhance the dynamic model of low Earth orbit spacecraft, further improving the accuracy of precise orbit determination and orbit prediction.

[0037] (2) In the method of this invention, each test particle can represent a large number of real gas molecules; in the simulation, the simulated particles are generated sequentially rather than simultaneously and are added to the computational domain one by one until the number of samples is sufficient to simulate the characteristics of the real physical process. The established atmospheric drag model contains the details and corresponding characteristics of atmospheric drag variation and has high fidelity.

[0038] (3) The present invention adopts the method of adding simulated particles one by one, which will not consume too much computing time and storage space, and is more suitable for the problem of multiple surface reflection flow caused by complex boundaries.

[0039] (4) The present invention adopts the KD tree structure to organize the three-dimensional model of the spacecraft, which greatly improves the search efficiency in the Monte Carlo simulation; the atmospheric drag modeling process in each direction can be combined with high-performance GPU for parallel calculation, which greatly shortens the modeling calculation time and improves the calculation efficiency.

[0040] (5) The present invention uses a grid file to store the atmospheric drag modeling results, which makes it convenient for users to obtain the atmospheric drag acceleration at any position and at any attitude by interpolation according to the satellite's attitude. Attached Figure Description

[0041] Figure 1 This is a basic schematic diagram of the present invention;

[0042] Figure 2 This is a flowchart of the present invention;

[0043] Figure 3 This is a schematic diagram for calculating atmospheric drag using physical simulation within the computational domain. Detailed Implementation

[0044] The following will be based on embodiments of the present invention. Figures 1-3The technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0045] To address the shortcomings of existing technologies, this invention proposes a versatile, high-fidelity, and high-precision method for modeling atmospheric drag on spacecraft. This method utilizes the Test Particle Monte Carlo (TPMC) method to simulate the interaction between atmospheric molecules or particles and the satellite surface, thereby calculating atmospheric drag. It is applicable to spacecraft of arbitrary shapes in sparse atmospheric environments and has broad applicability. Each test particle can represent a large number of real gas molecules, and particles can be added one by one during the simulation until the sampling quantity is sufficient to simulate the characteristics of real physical processes. The established atmospheric drag model contains detailed information and corresponding characteristics of atmospheric drag variations, exhibiting high fidelity. By introducing a KD-tree structure to organize the complex geometric model of the spacecraft, the search and computation efficiency of the algorithm is improved, and the parallel computing capabilities of high-performance GPUs significantly shorten the modeling computation time. The output data format is a grid file. Users can obtain atmospheric drag acceleration at any position and attitude by performing bilinear interpolation based on the satellite's attitude using the grid file.

[0046] refer to Figure 1 , Figure 2 The technical solution adopted in this invention uses data such as the satellite's geometry, surface material properties, position, velocity, atmospheric density, and space environment parameters as input. By setting the computational domain and state update conditions, the test particle Monte Carlo method is used for iteration. After stabilization, atmospheric drag model data is output in the form of a grid file. The steps are as follows:

[0047] The first step is to construct a 3D model database for spacecraft: the basic geometric prototype of the spacecraft is described in a parametric form, and complex 3D models of the spacecraft are constructed by combining the basic geometric prototypes; the 3D model database stores the parameters and surface material properties of each component of the 3D model of the spacecraft.

[0048] The constructed 3D model database is reorganized using a KD-tree data structure to improve the efficiency of retrieval and access to various components during subsequent modeling processes. In addition, satellite 3D model files can be directly read in, converted in format, and then used for 3D model quality checks.

[0049] The second step is to initialize the computation domain and set the state update rules:

[0050] First, a computational domain centered on the spacecraft model is constructed;

[0051] Secondly, atmospheric density is calculated based on space environment parameters, and particle flow is simulated in the computational domain. The initial state (P, V) of the test particle flow is set according to Boltzmann distribution, thermal motion equations, etc.

[0052] Then, state update rules are set, which include at least three states of the test particles: no collision and still within the computational domain, no collision but flying out of the computational domain, and collision with the spacecraft surface.

[0053] The third step is to simulate the interaction between the test particles and the spacecraft surface: using a detailed 3D model of the spacecraft, the interaction between the test particles and the spacecraft surface is simulated using the Monte Carlo simulation method. After multiple iterations and stabilization, the product of the atmospheric drag coefficient and the spacecraft surface area, Cd·A, is output. Similarly, the interaction between the test particles and the spacecraft in other directions can be simulated using Monte Carlo simulation, and the product of the atmospheric drag coefficient and the spacecraft surface area in other directions is calculated and output. The grid resolution is set to obtain the total number of directions. If the grid resolution is 1°, there are 361×181 directions, which means 361×181 Monte Carlo simulations are required. The simulation process in different directions can be performed in parallel on a high-performance computer to improve modeling efficiency.

[0054] The fourth step is the user algorithm for the model: After completing the atmospheric drag calculation in all directions, the product of the atmospheric drag coefficient and the surface area of ​​the spacecraft at any location will be output in the form of a grid file in the spacecraft's coordinate system. After obtaining the grid file, the user can use interpolation to calculate the atmospheric drag acceleration at any location and in any attitude.

[0055] This invention addresses the problems of complex atmospheric drag modeling methods and their difficulty in extending their application to various spacecraft by proposing a highly applicable, high-fidelity, and high-precision method for spacecraft atmospheric drag modeling. This method uses geometric prototypes to refine the spacecraft model and standardizes the database format for spacecraft atmospheric drag modeling, facilitating unified management. It utilizes the Monte Carlo simulation method with test particles to simulate the interaction between test particles and the spacecraft surface, making it suitable for spacecraft of arbitrary shapes in sparse atmospheric environments, thus demonstrating strong applicability. In the simulation, each test particle can represent a large number of real gas molecules, and the established model contains detailed information and corresponding characteristics of atmospheric drag variations, exhibiting high fidelity. Test particles are added to the computational domain one by one, minimizing computation time and storage space consumption, making it more suitable for complex boundary conditions leading to multiple surface reflection flows. A grid file is used to store the atmospheric drag modeling results, allowing users to easily obtain atmospheric drag acceleration at any position and attitude through interpolation. This method has broad application prospects in precise orbit determination of low Earth orbit spacecraft and high-precision satellite positioning.

[0056] Specifically, in the above four steps:

[0057] The specific steps for constructing the spacecraft 3D model database in the first step are as follows:

[0058] Step 1.1: In this invention, geometric prototypes such as planes, circles, spheres, rings, cones, cylinders, and frustums are set. Based on the actual shape of the satellite, geometric models of various components of the spacecraft are performed. The input parameters are the parameters of the corresponding geometric prototypes, and a complex and detailed three-dimensional model of the spacecraft is constructed by combining them.

[0059] Step 1.2: Using the constructed spacecraft 3D model database, the maximum geometric envelope of the spacecraft is used as the root node of the tree structure to construct a KD tree structure to store the spacecraft 3D model, which facilitates the retrieval and access efficiency of each component when simulating and testing the interaction between particles and the spacecraft surface.

[0060] Step 1.3: For spacecraft 3D model databases, satellite 3D model files can also be directly read in, converted to the correct format, and then the 3D model quality can be checked.

[0061] The specific steps for initializing the computation domain and setting the state update rules in the second step are as follows:

[0062] Step 2.1: Construct a computational domain centered on the aforementioned 3D model of the spacecraft and of appropriate size;

[0063] Step 2.2: Calculate the density and temperature of different types of atmospheric molecules based on the satellite's location, simulate the test particle flow in the computational domain, and set the initial position X0 and initial velocity V0 of the test particle flow.

[0064] Step 2.3: The test particle stream includes several test particles. The number of test particles follows an atmospheric density model, and the distribution follows a Boltzmann distribution. The initial position X0 can be calculated.

[0065] Step 2.4: Particle velocity consists of two parts: thermal velocity Vm and relative velocity Vr. The test particle obeys thermal motion, and the atmospheric inflow temperature T can be obtained through the atmospheric density model. ∞ The thermal velocity of the tested particle is then... In the formula, m is the mass of a single test particle. It is assumed that the spacecraft position remains unchanged and the test particle moves at a relative velocity Vr between the two.

[0066] Step 2.5: Set the time iteration step size Δt. Assume that the test particle moves in a straight line within a very short time iteration step size (e.g., 1ms). Update the position X0+1 and velocity V0+1 of all test particles, and determine whether a collision or flight out of the computation domain occurs.

[0067] Step 2.6: Assume that the collision between the test particle and the spacecraft surface can be considered an elastic collision;

[0068] If the test particle collides with the spacecraft surface, the velocity of the test particle after the collision is updated according to the laws of conservation of momentum and conservation of energy, and the rate of change of momentum Δp is recorded at this time.

[0069] If the test particle flies out of the computational domain without colliding with anything, then a new test particle is placed in the computational domain.

[0070] If the test particle does not collide with anything and does not fly out of the computational domain, the position and velocity of the test particle will remain P0+1 and V0+1.

[0071] The third step involves simulating the interaction between particles and the spacecraft surface. The specific steps are as follows:

[0072] Step 3.1: Using the Monte Carlo simulation method, simulate the interaction between the particle and the spacecraft surface, and record the rate of change of momentum Δp at the time of collision. i 3.2 After stabilization through multiple iterations, the atmospheric drag in this direction can be obtained. Then we can obtain the product C of the atmospheric drag coefficient and the spacecraft surface area. d A;

[0073] Step 3.3: Set the grid resolution to 1°, then there are 361×181 directions. Adjust the position of the spacecraft in the computational domain, perform Monte Carlo simulations in other directions according to the above process, and output the product of the atmospheric drag coefficient and the surface area of ​​the spacecraft in other directions.

[0074] Step 3.4: Simulation processes in different directions can be performed in parallel on high-performance computers to improve modeling efficiency; if computing performance allows, the grid resolution can be further reduced to obtain a more refined 3D model of the spacecraft.

[0075] In the fourth step, the specific steps of the model user algorithm are as follows:

[0076] Step 4.1 After completing the atmospheric drag modeling calculations in all directions, output the product of the atmospheric drag coefficient and the spacecraft surface area at any location in the spacecraft coordinate system as a grid file.

[0077] Step 4.2: After obtaining the grid file, the user can use interpolation to calculate the atmospheric drag acceleration at any position and in any attitude.

[0078] Compared with the prior art, the advantages of the present invention are as follows:

[0079] (1) This invention simulates the interaction between test particles and the spacecraft surface using the Monte Carlo simulation method for test particles. It is applicable to spacecraft of arbitrary shapes in sparse atmospheric environments, has strong applicability, and can be flexibly used for modeling different spacecraft. As a non-conservative force modeling method, it can improve and enhance the dynamic model of low Earth orbit spacecraft, further improving the accuracy of precise orbit determination and orbit prediction.

[0080] (2) In the method of this invention, each test particle can represent a large number of real gas molecules; in the simulation, the simulated particles are generated sequentially rather than simultaneously and are added to the computational domain one by one until the number of samples is sufficient to simulate the characteristics of the real physical process. The established atmospheric drag model contains the details and corresponding characteristics of atmospheric drag variation and has high fidelity.

[0081] (3) The present invention adopts the method of adding simulated particles one by one, which will not consume too much computing time and storage space, and is more suitable for the problem of multiple surface reflection flow caused by complex boundaries.

[0082] (4) The present invention adopts the KD tree structure to organize the three-dimensional model of the spacecraft, which greatly improves the search efficiency in the Monte Carlo simulation; the atmospheric drag modeling process in each direction can be combined with high-performance GPU for parallel calculation, which greatly shortens the modeling calculation time and improves the calculation efficiency.

[0083] (5) The present invention uses a grid file to store the atmospheric drag modeling results, which makes it convenient for users to obtain the atmospheric drag acceleration at any position and at any attitude by interpolation according to the satellite's attitude.

[0084] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Any modifications, alterations, substitutions, or variations made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention shall fall within the protection scope defined by the claims of the present invention.

Claims

1. A high-fidelity, high-precision atmospheric drag modeling method for spacecraft, characterized in that, Includes the following steps: The first step is to construct a 3D model database for the spacecraft: the basic geometric prototype of the spacecraft is described in a parametric form, and a 3D model of the spacecraft is constructed by combining the basic geometric prototype; the 3D model database stores the geometric parameters and surface material properties of each component of the 3D model of the spacecraft. The second step is to initialize the computational domain and set the state update rules: construct a computational domain centered on the three-dimensional model of the spacecraft; simulate the test particle flow within the computational domain and set the initial state of the test particle flow; The third step is to simulate and test the interaction between the particles and the spacecraft surface: using the Monte Carlo simulation method, the interaction between the particles and the spacecraft surface is simulated and tested. After multiple iterations and stabilization, the product of the atmospheric drag coefficient in each direction and the surface area of ​​the spacecraft is output. The fourth step is to calculate the atmospheric drag in all directions. Then, in the spacecraft coordinate system, output the product of the atmospheric drag coefficient and the surface area of ​​the spacecraft at any position. This allows us to calculate the atmospheric drag acceleration of the spacecraft's three-dimensional model at any position and in any attitude. The second step, setting the initial state of the test particle flow, includes the following steps: Simulate the test particle flow within the computational domain and set the initial position X0 and initial velocity V0 of the test particle flow; The test particle stream consists of several test particles, the number of which follows an atmospheric density model and the distribution follows a Boltzmann distribution. The initial position X0 is calculated. The initial velocity of the test particle stream includes the thermal velocity V. m and relative velocity V r ; If the test particle flow obeys thermal motion, the atmospheric inflow temperature can be obtained through an atmospheric density model. The test particle's thermal velocity is: In the formula, m is the mass of a single test particle; If the spacecraft's position remains unchanged, the test particle moves at a relative velocity V. r sports; Set the time iteration step Δt Assuming the test particles move in a straight line within the iteration step, update the positions X of all test particles. 0+1 and speed V 0+1 And determine whether a collision or flight out of the computational domain has occurred; If the test particle collides with the spacecraft surface, update the particle's velocity after the collision and record the rate of change of momentum at that moment. ; If the test particle flies out of the computational domain without colliding with anything, then a new test particle is placed in the computational domain. If the test particle does not collide with anything and does not fly out of the computational domain, then the position and velocity of the test particle remain X. 0+1 and V 0+1； The third step, which uses Monte Carlo simulation to simulate the interaction between particles and the spacecraft surface, includes the following steps: The Monte Carlo simulation method was used to simulate the interaction between particles and the spacecraft surface, and the rate of change of momentum at the time of collision was recorded. ; After multiple iterations and stabilization, the atmospheric drag in this direction was obtained. In the formula Δt The time iteration step is given, which is used to obtain the product of the atmospheric drag coefficient and the spacecraft surface area. ; Set the grid resolution to get the total number of directions, and adjust the spacecraft's position in the computational domain; Repeat the above steps to perform Monte Carlo simulations in other directions and output the product of atmospheric drag coefficients and spacecraft surface area in those other directions.

2. The high-fidelity, high-precision atmospheric drag modeling method for spacecraft according to claim 1, characterized in that, In the first step, geometric models of each component of the spacecraft are performed based on the actual shape of the satellite. The constructed three-dimensional model database of the spacecraft is then reorganized using a KD-tree data structure.

3. The high-fidelity, high-precision atmospheric drag modeling method for spacecraft according to claim 1, characterized in that: The second step also includes setting state update rules, which include at least three states of test particles: no collision and still within the computational domain, no collision but flying out of the computational domain, and colliding with the surface of the spacecraft.

4. The high-fidelity, high-precision atmospheric drag modeling method for spacecraft according to claim 1, characterized in that: In the fourth step, the product of the atmospheric drag coefficient and the spacecraft surface area at any location is output in the form of a grid file. After obtaining the grid file, the user can use interpolation to calculate the atmospheric drag acceleration at any location and in any attitude.

Images