Multi-body separation rapid simulation method based on unsteady aerodynamic modeling

Abstract

The invention provides a multi-body separation rapid simulation method based on unsteady aerodynamic modeling. The method comprises the steps that firstly, a static aerodynamic force database and unsteady aerodynamic force data of a single typical state are obtained through CFD calculation; and then the two types of data are fused and modeled, and a high-precision agent model capable of predicting aerodynamic force in the dynamic process is constructed. During simulation, the real-time position and posture of a separated body are input into the model, the current aerodynamic force is quickly predicted, the motion state is updated by solving a rigid body six-degree-of-freedom motion equation, surface grid motion is driven to achieve visualization, and loop iteration is conducted till separation is finished. According to the method, static and unsteady aerodynamic forces are fused, the simulation efficiency is improved by several orders of magnitude compared with a traditional method while the precision of capturing the dynamic aerodynamic effect is guaranteed, and the method can be widely applied to design and analysis in the aerospace and weapon engineering fields such as aircraft separation and ammunition scattering.

Description

Technical Field

[0001] This invention relates to the field of multibody separation simulation technology, and in particular to a fast multibody separation simulation method based on unsteady aerodynamic modeling. Background Technology

[0002] In aerospace, weaponry, and other engineering fields, the multi-body separation process is a critical factor affecting equipment safety and mission reliability. For example, when a missile is launched from an aircraft, if the two stages collide or lose control of their separation attitude during the separation process, it will directly lead to mission failure or even equipment damage. During the launch and orbit insertion of a multi-stage launch vehicle, the safety of the two-stage separation determines the success or failure of the launch mission. Therefore, accurate and efficient simulation analysis of the multi-body separation process is crucial during the equipment design phase.

[0003] Currently, multibody separation simulation mainly employs two technical approaches: one is a simplified simulation method based on empirical formulas or static aerodynamic models. This method obtains static aerodynamic data of the separated bodies through experiments, establishes empirical formulas or lookup tables, and then combines them with rigid body motion equations to solve for the separation trajectory. Its advantage lies in its high computational efficiency, but it has significant drawbacks: it ignores the unsteady effects of airflow during the separation process (such as the generation and shedding of separation vortices, airflow disturbances, etc.), resulting in large aerodynamic prediction errors. Especially in scenarios with severe airflow disturbances in the early stages of separation, the simulation results can deviate from the actual situation by more than 20%, failing to meet the requirements of high-precision design.

[0004] Another approach is a high-precision simulation method based on unsteady computational fluid dynamics (UCFD). This method directly solves for the unsteady flow field during the separation process numerically, accurately capturing unsteady aerodynamic characteristics. However, since unsteady flow field calculations require significant computational resources and time, for complex multibody separation processes (such as sequential separation of multiple loads or complex-shaped separation bodies), a single simulation often takes several days or even weeks, resulting in extremely low computational efficiency and making it difficult to meet the needs of rapid iterative equipment design.

[0005] Therefore, how to improve the efficiency of multibody separation simulation while ensuring simulation accuracy, and how to balance the contradiction between accuracy and efficiency, has become an urgent technical problem to be solved in the field of multibody separation simulation technology. Summary of the Invention

[0006] The purpose of this invention is to provide a fast simulation method for multibody separation based on unsteady aerodynamic modeling. By fusing static and unsteady aerodynamic data, a high-precision prediction model is constructed, which significantly improves computational efficiency while ensuring simulation accuracy.

[0007] According to the purpose of this invention, a fast simulation method for multibody separation based on unsteady aerodynamic modeling is provided, comprising the following steps: S1: Static aerodynamic data of the separated body under different states are obtained by computational fluid dynamics (CFD) method to construct a static aerodynamic database; at the same time, a single typical motion state of the separated body is selected, and the unsteady motion process under this state is calculated by CFD method to obtain unsteady aerodynamic data. S2: The static and unsteady aerodynamic data obtained in step S1 are fused and modeled to construct an aerodynamic prediction model. This model can predict the aerodynamic forces at the corresponding time or position during the separation process based on the position and attitude information of the separated body. S3: Set the initial time t0 for the multibody separation simulation, obtain the position and attitude information of the separated bodies at time t0, and input them into the aerodynamic prediction model constructed in step S2 to obtain the aerodynamic force at time t0. S4: Substitute the aerodynamic force at time t0 obtained in step S3 into the rigid body's six-degree-of-freedom motion equation, and obtain the position and attitude information of the separated body at the next time t1 through numerical solution, where t1 = t0 + Δt, and Δt is the set time step. S5: Based on the position and attitude information of the separated body at time t1 obtained in step S4, the surface mesh of the separated body is moved and rotated to make the surface mesh consistent with the actual position and attitude of the separated body, thereby realizing the visualization of the separation process; S6: Determine whether the set separation time has been reached. If not, use the position and attitude information at time t1 as the new input and repeat steps S3 to S5. If the time has been reached, stop the simulation and output the separation trajectory data and attitude change data of the separated body during the entire separation process.

[0008] Furthermore, in step S1, the different states include different angles of attack, sideslip angles, Mach numbers, and flight altitudes of the separated body; the static aerodynamic data include axial force, normal force, lateral force, roll moment, yaw moment, and pitch moment.

[0009] Furthermore, in step S1, when calculating the unsteady motion process under a single typical motion state, an unsteady CFD solution method is adopted, and the solution time step is set according to the actual calculation state to ensure the time resolution of the unsteady aerodynamic data.

[0010] Furthermore, in step S2, the data fusion modeling adopts a weighted fusion algorithm. The weight coefficients of static aerodynamic data and unsteady aerodynamic data are adaptively adjusted according to the dynamic characteristics of the separation process. The dynamic characteristics include the motion acceleration, angular velocity and airflow disturbance intensity of the separated body.

[0011] Furthermore, in step S2, the data fusion modeling adopts the Co-Kriging model, treating static aerodynamic data as low-fidelity data and unsteady aerodynamic data as high-fidelity data. The model hyperparameters are solved by optimization algorithms to construct an aerodynamic prediction model.

[0012] Further, in step S4, the equations of motion for the rigid body with six degrees of freedom are: Translation: ; Rotation: ; The above equations of motion are solved using second-order Newmark integrals: ; in, χ represent v or ω, In practical solutions, the rotation equations are represented using quaternions to avoid singularities caused by gimbal lock-up.

[0013] Furthermore, in step S4, the rotation equation is expressed in quaternions to avoid singularities caused by gimbal lock.

[0014] Furthermore, in step S5, when moving and rotating the surface mesh, the displacement of each node is calculated based on the position and attitude changes of the separated body, thereby realizing mesh adjustment.

[0015] Furthermore, in step S6, the condition for determining whether the separation process has ended is: the separation time reaches the preset time or the distance between the separated bodies is greater than the preset safe distance.

[0016] Furthermore, the preset safety distance is determined based on the size of the separated body, and is 3 to 5 times the maximum characteristic size of the separated body.

[0017] This invention's technical solution constructs a highly efficient prediction model by fusing static and unsteady aerodynamic data, significantly improving simulation efficiency while maintaining accuracy. Compared to traditional tightly coupled CFD / 6-DOF methods, computational efficiency can be improved by two to three orders of magnitude, supporting large-scale parametric studies and rapid iterations during the design phase. Simultaneously, the model effectively incorporates unsteady aerodynamic effects, significantly improving the accuracy of separation trajectory and attitude prediction, overcoming the shortcomings of traditional engineering methods that neglect dynamic aerodynamics. Furthermore, this method is highly versatile and applicable to a wide range of scenarios, and can be widely used in various engineering fields such as aircraft external stores separation, rocket stage separation, munitions disposal, and multibody motion disturbance analysis. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the method of an embodiment of the present invention; Figure 2 This is a schematic diagram of the external shape of an embodiment of the present invention, showing a missile launcher mounted on a wing. Figure 3 This is a trajectory prediction effect of a specific implementation example of the method in the embodiments of the present invention; Figure 4 This is the posture prediction effect of a specific implementation case of the method in the embodiments of the present invention; Figure 5 This is a visual illustration of the separation process in a specific implementation example of the method of this invention. Detailed Implementation

[0020] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0022] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0023] Example 1 like Figures 1-5 As shown, a fast simulation method for multibody separation based on unsteady aerodynamic modeling includes the following steps: S1: Obtain static aerodynamic data of the separated body under different states by computational fluid dynamics (CFD) method and construct a static aerodynamic database; at the same time, select a single typical motion state of the separated body and calculate the unsteady motion process under this state by CFD method to obtain unsteady aerodynamic data. S2: The static and unsteady aerodynamic data obtained in step S1 are fused and modeled to construct an aerodynamic prediction model. This model can predict the aerodynamic forces at the corresponding time or position during the separation process based on the position and attitude information of the separated body. S3: Set the initial time t0 for the multibody separation simulation, obtain the position and attitude information of the separated bodies at time t0, and input them into the aerodynamic prediction model constructed in step S2 to obtain the aerodynamic force at time t0. S4: Substitute the aerodynamic force at time t0 obtained in step S3 into the rigid body six-degree-of-freedom (6DOF) motion equation, and obtain the position and attitude information of the separated body at the next time t1 through numerical solution, where t1=t0+Δt, and Δt is the set time step; S5: Based on the position and attitude information of the separated body at time t1 obtained in step S4, the surface mesh of the separated body is moved and rotated to make the surface mesh consistent with the actual position and attitude of the separated body, thereby realizing the visualization of the separation process; S6: Determine whether the set separation time has been reached. If not, use the position and attitude information at time t1 as the new input and repeat steps S3 to S5. If the time has been reached, stop the simulation and output the separation trajectory data and attitude change data of the separated body during the entire separation process.

[0024] Specifically, in step S1, the "different states" include different angles of attack, sideslip angles, Mach numbers, and flight altitudes of the separated body; the static aerodynamic data include axial force, normal force, lateral force, roll moment, yaw moment, and pitch moment.

[0025] In step S1, when calculating the unsteady motion process under a single typical motion state, an unsteady CFD solution method is used, and the solution time step is set according to the actual calculation state to ensure the time resolution of the unsteady aerodynamic data.

[0026] In step S2, the data fusion modeling adopts a weighted fusion algorithm. The weight coefficients of static aerodynamic data and unsteady aerodynamic data are adaptively adjusted according to the dynamic characteristics of the separation process. The dynamic characteristics include the motion acceleration, angular velocity and airflow disturbance intensity of the separated body. After training, the aerodynamic prediction model is obtained.

[0027] In step S2, the data fusion modeling adopts the Co-Kriging model, which treats static aerodynamic data as low-fidelity data and unsteady aerodynamic data as high-fidelity data. The model hyperparameters are solved by optimization algorithm to construct an aerodynamic prediction model.

[0028] In step S4, the equations of motion for the rigid body with six degrees of freedom are: Translation: ; Rotation: ; The above equations of motion are solved using second-order Newmark integrals: ; in, χ represent v or ω, In practical solutions, the rotation equations are represented using quaternions to avoid singularities caused by gimbal lock-up.

[0029] In step S5, when moving and rotating the surface mesh, the displacement of each node is calculated based on the position and attitude changes of the separated body, thereby achieving mesh adjustment.

[0030] In step S6, the condition for determining whether the separation process has ended is: the separation time reaches the preset time or the distance between the separated bodies is greater than the preset safety distance; the preset safety distance is determined according to the size of the separated bodies and is 3 to 5 times the maximum feature size of the separated bodies.

[0031] Example 2 like Figures 1-5As shown in the figure, this embodiment presents a rapid prediction method for multibody separation based on unsteady aerodynamic modeling, comprising: obtaining high-precision steady aerodynamic forces and high-precision unsteady aerodynamic forces of the separated bodies; modeling the steady and unsteady aerodynamic forces using a fusion modeling method; inputting the initial position and attitude of the target state separated bodies into the trained aerodynamic prediction model to obtain the aerodynamic forces / torques acting on the separated bodies at the current time step; inputting the aerodynamic forces / torques of the separated bodies at the current time step into the six-degree-of-freedom solution module to obtain the displacement and attitude change information of the separated bodies at the next time step; inputting the attitude and displacement change information of the splitting fluids into the mesh motion module to output a new surface mesh after translation and rotation; inputting the new position and attitude information of the separated bodies into the aerodynamic prediction model, and repeating the above process until the preset separation time is reached or the distance between the separated bodies is greater than the safe separation distance.

[0032] The aforementioned rapid prediction method for multibody separation based on unsteady aerodynamic modeling employs computational fluid dynamics to obtain high-precision steady / unsteady aerodynamic information. Specifically, the numerical algorithm uses a lattice-centered unstructured finite volume method to solve the Navier-Stokes equations, the Roe scheme to calculate inviscid flux, and the Green-Gauss theorem to calculate the gradient distribution within the cell to obtain second-order spatial accuracy. A Venkatakrishnan limiter suppresses overshoot and oscillations near discontinuities, and the viscous flux is calculated using a central difference scheme. The separation motion calculation also needs to consider the grid velocity and the acceleration of the solid boundary. The SST turbulence model is selected. k - ω Turbulence model. All simulations are unsteady and calculated using a two-time-step method. The physical time layer uses second-order backward difference, and the pseudo-time layer uses LU-SGS implicit time advancement.

[0033] In the aforementioned rapid prediction method for multibody separation based on unsteady aerodynamic modeling, the steady / unsteady aerodynamic fusion model employs the Co-Kriging model as the data fusion algorithm. This aims to utilize a small amount of high-fidelity, high-cost data to "calibrate" or "correct" a large amount of low-cost, low-fidelity data, thereby constructing an efficient and accurate prediction model. Co-Kriging is a multi-fidelity modeling method that treats the static aerodynamic database obtained based on CFD as low-fidelity (LF) data and the unsteady aerodynamic time-domain data obtained in the previous steps as high-fidelity (HF) data. The basic idea of ​​this model is that the output of the high-fidelity model can be expressed as the sum of the low-fidelity model output and a correction term (residual). The specific process of model construction is as follows: First, all input parameters and output data are normalized to improve the numerical stability of the model; then, optimization algorithms such as maximum likelihood estimation (MLE) are used to solve for the hyperparameters in the Co-Kriging model, such as the coefficient ρ describing the correlation between high-fidelity and low-fidelity data and the parameters of the Gaussian process covariance function. The optimized Co-Kriging model can quickly predict aerodynamic forces and torques that take into account unsteady effects based on the real-time motion state (position, attitude, angular velocity, etc.) of the separated body.

[0034] In the above-mentioned fast prediction method for multibody separation based on unsteady aerodynamic modeling, the iterative solution steps for separation trajectory and attitude are the core loop for simulating the entire separation process. The specific operation is as follows: First, the initial conditions for separation are set, including the initial relative position, attitude, linear velocity, and angular velocity of the separated objects (usually determined by the deployment mechanism). Then, a time step Δ is initiated. t The cycle of advancement: 1) At the current moment t 0. Taking the initial relative position and attitude of the separated bodies as input, the constructed Co-Kriging fusion proxy model is invoked to quickly obtain the aerodynamic vectors acting on the separated bodies. F a and aerodynamic torque vector M a .

[0035] 2) F a and M a Along with any additional forces applied (such as gravity and ejection force), these are substituted into the six-degree-of-freedom rigid body motion equations describing the object's spatial motion, and solved using second-order Newmark integration. The solution yields the next moment's motion. t 1= t 0+Δ t The new motion state (position, attitude, velocity, angular velocity) of the separated body.

[0036] 4) To achieve process visualization and subsequent possible flow field analysis, based on t The pose information at time 1 is used to perform rigid translation and rotation transformations on the three-dimensional surface mesh model of the separated body.

[0037] 5) Finally, check whether the separation termination criteria are met, i.e., whether the total simulation time has reached the set upper limit or whether the centroid distance between the separated bodies has exceeded the preset safety threshold. If not, update the current time to... t 1. Return to step 1) and begin the next round of iterative calculation. If the conditions are met, the loop ends, and the complete separation trajectory, attitude change history, and other data are output.

[0038] Example 3 like Figures 1-5 As shown in this embodiment, a fast multibody separation simulation method based on unsteady aerodynamic modeling is proposed. The method includes: First, a static aerodynamic database is obtained using computational fluid dynamics (CFD) methods, and then the unsteady motion process under a single state is calculated to obtain unsteady aerodynamic data. Next, static aerodynamic data and unsteady aerodynamic data are fused and modeled to construct a model that can accurately predict aerodynamic forces at different times or locations during the separation process. Subsequently, based on this model, the corresponding aerodynamic forces are obtained by using the position and attitude information of the separated body at the initial moment, and then substituted into the rigid body motion equation to solve for the position and attitude information of the separated body at the next moment. Simultaneously, the surface mesh of the separated body is moved and rotated based on the current position and attitude information to visualize the separation process; Finally, repeat the above steps of aerodynamic prediction, motion state solution, and mesh adjustment until the multibody separation process is completed.

[0039] Specifically, taking wing-mounted missile deployment as an example, the method includes the following steps: Step 1: Basic Parameter Setting. Based on the target case, determine the basic physical properties of the separated body, such as its geometry, mass, and moment of inertia matrix; set the environmental parameters associated with the initial motion state (such as flight altitude, Mach number, etc.) to provide basic input for subsequent aerodynamic data acquisition and simulation calculations.

[0040] Step Two: Static Aerodynamic Database Construction. Based on the typical separation conditions in Step One, determine the approximate motion state of the separated body, such as angle of attack, sideslip angle, and distance traveled, ensuring that the conditions cover the entire separation process. Use a self-developed CFD solver or general commercial software to perform static aerodynamic calculations on the separated body. After convergence, extract static aerodynamic data such as lift, drag, side force, rolling moment, yaw moment, and pitching moment, organize and store them in a unified data format, and construct a static aerodynamic database.

[0041] Step 3: Acquisition of Unsteady Aerodynamic Data. This step is conducted concurrently with Step 2. Based on the dynamic characteristics analysis of the separation process, a single typical motion state is selected. Unsteady aerodynamic calculations of the separated body are performed using a self-developed CFD solver or general commercial software to obtain complete unsteady aerodynamic variation patterns. Unsteady aerodynamic data for each time step are extracted to form a time-series data set.

[0042] Step 4: Sample Integration and Preprocessing. The static aerodynamic data (as the base sample) and unsteady aerodynamic data (as the dynamic correction sample) obtained in Steps 2 and 3 are integrated to form the original sample set of the aerodynamic prediction model. At the same time, the input parameters (such as operating condition parameters and motion state parameters) and output parameters (aerodynamic data) of all samples are normalized to eliminate the influence of parameter magnitude differences on model training.

[0043] Step 5: Sample set partitioning. Using the sample partitioning strategy commonly used in cross-validation, the original sample set obtained in Step 4 is divided into a training set (for model parameter learning), a validation set (for parameter adjustment and overfitting determination during model training), and a test set (for final model performance evaluation) in an 8:1:1 ratio.

[0044] Step Six: Model Training. This example uses the Co-Kriging model as the data fusion algorithm for steady / unsteady aerodynamic fusion modeling. The static aerodynamic database obtained in Step Five is considered low-fidelity (LF) data, and the unsteady aerodynamic time-domain data is considered high-fidelity (HF) data. The hyperparameters in the Co-Kriging model are solved using the maximum likelihood estimation (MLE) optimization algorithm. The coefficient ρ describing the correlation between high-fidelity and low-fidelity data, as well as the parameters of the Gaussian process covariance function, are obtained. This enables rapid prediction of aerodynamic forces and moments considering unsteady effects based on the real-time motion state (position, attitude, angular velocity, etc.) of the separated body.

[0045] Step 7: Initial aerodynamic force prediction. Set the initial time t0 for the multibody separation simulation, and obtain the position information (such as centroid coordinates) and attitude information (such as roll angle, yaw angle, and pitch angle) of the separated body at this time. Input this information along with the corresponding operating parameters into the aerodynamic force prediction model constructed in Step 6. The model outputs the aerodynamic force vector and aerodynamic moment vector of the separated body at time t0.

[0046] Step 8: Solving for the motion state at the next moment. The aerodynamic force vector and aerodynamic moment vector output from Step 7 are used as input to the rigid body motion equation solving module. Combined with the physical parameters of the separated body set in Step 1, such as the mass and moment of inertia matrix, as well as environmental load parameters such as the gravity vector, a second-order Newmark integral is used to solve the problem. After solving, the new motion state (position, attitude, velocity, angular velocity) of the separated body at time t1 is obtained, providing input for subsequent mesh adjustment and cyclic simulation.

[0047] Step Nine: Mesh Motion and Visualization. Based on the motion state of the separated body at time t1 obtained in Step Eight, this state information is input into the dynamic mesh solver module. This module pre-stores the initial surface mesh data of the separated body. According to the position offset and attitude rotation angle of the separated body at time t1, the surface mesh is moved and rotated and adjusted, and a standard plt format file is output. Using Tecplot post-processing software, the surface mesh data of each time step output by the dynamic mesh solver module is read, and combined with the motion state information output by the rigid body motion equation solver module, a time series animation of the separation process is generated.

[0048] Step 10: Separation End Determination. Based on the safety requirements of the separation scenario, set the determination conditions for the end of separation: including whether the centroid distance between the separated bodies and the shortest distance on the surface reach the preset safety distance (the safety distance is set according to the geometric dimensions of the separated bodies, for example, 3 to 5 times the maximum feature size of the separated bodies), and whether the velocity change rate of the separated bodies meets the motion stability requirements; if all determination conditions are met, the separation process is determined to be over; if not, time t1 is taken as the new time t0, and the process returns to Step 7 to repeat the cycle of aerodynamic prediction, motion state solution, mesh adjustment and visualization.

[0049] This invention improves the overall simulation efficiency by two to three orders of magnitude compared to the traditional tightly coupled CFD / 6-DOF method through one-time offline data preparation (static database and a small amount of non-steady computation) and model building, making it possible to conduct large-scale parametric studies and optimizations during the design phase. This invention introduces a small amount of high-fidelity unsteady CFD calculation data and uses a data fusion model to effectively fuse it with static data, so that the final surrogate model can accurately capture key unsteady physical phenomena such as aerodynamic delay and dynamic damping during the separation process, thereby ensuring the accuracy of separation trajectory and attitude prediction. The method framework of this invention is universal and applicable not only to the separation of aircraft from external stores, but also to various multibody dynamics problems involving complex aerodynamic disturbances, such as rocket stage separation, probe landing, and multi-UAV formation flight.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fast simulation method for multibody separation based on unsteady aerodynamic modeling, characterized in that, Includes the following steps: S1: Static aerodynamic data of the separated body under different states are obtained by computational fluid dynamics (CFD) method to construct a static aerodynamic database; at the same time, a single typical motion state of the separated body is selected, and the unsteady motion process under this state is calculated by CFD method to obtain unsteady aerodynamic data. S2: The static and unsteady aerodynamic data obtained in step S1 are fused and modeled to construct an aerodynamic prediction model. This model can predict the aerodynamic forces at the corresponding time or position during the separation process based on the position and attitude information of the separated body. S3: Set the initial time t0 for the multibody separation simulation, obtain the position and attitude information of the separated bodies at time t0, and input them into the aerodynamic prediction model constructed in step S2 to obtain the aerodynamic force at time t0. S4: Substitute the aerodynamic force at time t0 obtained in step S3 into the rigid body's six-degree-of-freedom motion equation, and obtain the position and attitude information of the separated body at the next time t1 through numerical solution, where t1 = t0 + Δt, and Δt is the set time step. S5: Based on the position and attitude information of the separated body at time t1 obtained in step S4, the surface mesh of the separated body is moved and rotated to make the surface mesh consistent with the actual position and attitude of the separated body, thereby realizing the visualization of the separation process; S6: Determine whether the set separation time has been reached. If not, use the position and attitude information at time t1 as the new input and repeat steps S3 to S5. If the time has been reached, stop the simulation and output the separation trajectory data and attitude change data of the separated body during the entire separation process.

2. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S1, the different states include different angles of attack, sideslip angles, Mach numbers, and flight altitudes of the separated body; the static aerodynamic data include axial force, normal force, lateral force, roll moment, yaw moment, and pitch moment.

3. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S1, when calculating the unsteady motion process under a single typical motion state, an unsteady CFD solution method is used, and the solution time step is set according to the actual calculation state to ensure the time resolution of the unsteady aerodynamic data.

4. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S2, the data fusion modeling adopts a weighted fusion algorithm. The weight coefficients of static aerodynamic data and unsteady aerodynamic data are adaptively adjusted according to the dynamic characteristics of the separation process. The dynamic characteristics include the motion acceleration, angular velocity and airflow disturbance intensity of the separated body.

5. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S2, the data fusion modeling adopts the Co-Kriging model, which treats static aerodynamic data as low-fidelity data and unsteady aerodynamic data as high-fidelity data. The model hyperparameters are solved by optimization algorithm to construct an aerodynamic prediction model.

6. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S4, the equations of motion for the rigid body with six degrees of freedom are: Translation: ; Rotation: ; The above equations of motion are solved using second-order Newmark integrals: ； in, χ represent v or ω, In practical solutions, the rotation equations are represented using quaternions to avoid singularities caused by gimbal lock-up.

7. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 6, characterized in that, In step S4, the rotation equation is expressed in quaternions to avoid singularities caused by gimbal lock.

8. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S5, when moving and rotating the surface mesh, the displacement of each node is calculated based on the position and attitude changes of the separated body, thereby achieving mesh adjustment.

9. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 1, characterized in that, In step S6, the condition for determining whether the separation process has ended is: the separation time reaches the preset time or the distance between the separated bodies is greater than the preset safe distance.

10. The fast simulation method for multibody separation based on unsteady aerodynamic modeling according to claim 9, characterized in that, The preset safety distance is determined based on the size of the separated body, and is 3 to 5 times the maximum characteristic size of the separated body.

Images