Intelligent optimization method for multi-satellite separation spring parameters of stacked satellites
By establishing a separation dynamics model and optimizing the separation spring parameters using a generalized reduction gradient algorithm, the problem of selecting separation spring parameters for stacked satellites was solved, achieving efficient spring parameter optimization and safety assessment, and reducing the risk of multi-satellite separation collisions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-21
- Publication Date
- 2026-04-14
AI Technical Summary
Traditional methods struggle to effectively select the spring parameters for separating multiple stacked satellites, leading to a high risk of collisions during satellite separation. Existing technologies also struggle to achieve efficient spring parameter optimization.
A separation dynamics model was established using multibody dynamics simulation analysis software. The stiffness parameterization model of the separation spring was used, and iterative optimization was performed using the multidisciplinary optimization software Isight and the generalized reduction gradient algorithm. The objective function for multi-star separation safety assessment and the constraints of optimization parameters were established, and parameter optimization was achieved through the FMI interface.
The method of quickly establishing a multi-planet separation spring parameter optimization model reduces the difficulty of script writing for multi-planet separation simulation models, improves the efficiency of spring parameter selection, and reduces the risk of multi-planet separation collisions.
Smart Images
Figure CN121859513A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of satellite-rocket separation safety analysis, and relates to an intelligent optimization method for the separation spring parameters of stacked satellites. Background Technology
[0002] The layered stacking configuration of flat-panel satellites is an important launch method for multiple satellites on a single rocket. By utilizing the pressure difference of the separation springs between the satellites, satellites in different layers can achieve a slight velocity difference, reducing the risk of collisions during separation. Traditional trial-and-error and enumeration methods are insufficient to obtain multiple sets of separation spring parameters. Summary of the Invention
[0003] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose an intelligent optimization method for the separation spring parameters of stacked satellites. The method uses an intelligent optimization algorithm to solve the problem of selecting inter-satellite separation springs for stacked satellites, and has high generalization ability, which can be adapted to the selection and calculation of vertical separation spring parameters for various multi-satellite stacking.
[0004] The solution of the present invention is:
[0005] A method for intelligent optimization of spring parameters for multi-satellite separation in stacked satellites includes:
[0006] Step 1: Establish a dynamic simulation model of inter-satellite separation between stacked satellites using multibody dynamics simulation analysis software;
[0007] Step 2: Establish a parameterized model of the stiffness of the separation spring in the separation dynamics simulation model, and use the parameterized model of the stiffness of the separation spring as the design variable;
[0008] Step 3: In the parameterized model of the separation spring stiffness, establish the objective function for multi-star separation safety assessment and the optimization parameter constraints;
[0009] Step 4: Based on the objective function and optimization parameter constraints of the multi-satellite separation safety assessment in Step 3, select the separation spring optimization algorithm and iteratively optimize the stiffness parameterization model of the separation spring in Step 2.
[0010] Step 5: Complete the final optimization of the inter-satellite separation dynamics simulation model among stacked satellites.
[0011] In the above-mentioned intelligent optimization method for the separation spring parameters of stacked satellites, in step one, the multibody dynamics software reads the satellite 3D model entity in intermediate format from the 3D modeling software to establish a simulation model of inter-satellite separation dynamics between stacked satellites.
[0012] In the above-mentioned intelligent optimization method for the separation spring parameters of stacked satellites, the method for establishing the stiffness parameterization model of the separation spring in step two is as follows:
[0013] The mechanical properties of the release spring are simulated using a triaxial force approach, and the equation is as follows:
[0014]
[0015] In the formula, X i The spatial position of the free end of the separation direction spring before the separation of the i-th layer satellite;
[0016] X i+1 The coordinates of the point of application of the separation spring thrust before the separation of satellites in the i-th and (i+1)-th layers; X before multi-satellite separation i =X i+1 ;
[0017] L is the maximum length that the free end of the release spring can extend;
[0018] F i The separation spring pressure between the i-th layer satellite and the (i+1)-th layer satellite;
[0019] K i Let the spring stiffness be the separation stiffness of the i-th layer satellite;
[0020] L0 is the compression length of the release spring.
[0021] In the aforementioned intelligent optimization method for multi-satellite separation spring parameters of stacked satellites, the objective function for evaluating the safety of multi-satellite separation in step three is:
[0022]
[0023] In the formula, Obj is the target value;
[0024] n is the total number of stacked satellite layers that are docked with the launch vehicle separation mechanism;
[0025] V i Let be the relative velocity between the i-th layer satellite and the launch vehicle separation mechanism;
[0026] The objective function for assessing the safety of multi-satellite separation is the sum of squares of the difference in separation velocities between two adjacent satellite layers minus 100 mm / s.
[0027] In the above-mentioned intelligent optimization method for the spring parameters of multi-satellite stacking separation, the constraint condition for the optimization parameters is:
[0028] Among them, the lowest layer of satellites, which is in direct contact with the launch vehicle separation mechanism, separates from the mechanism at a speed greater than 200 mm / s; and the separation spring stiffness of the i-th layer of satellites is greater than 0, i.e., K i>0.
[0029] In the above-mentioned intelligent optimization method for the separation spring parameters of stacked satellites, in step four, the multidisciplinary optimization software Isight and the multibody dynamics simulation analysis software are integrated using the FMI interface. The separation spring optimization parameters and objective function established in the multibody dynamics software are quickly transferred to the multidisciplinary optimization software Isight, and combined with the optimization algorithm function package in the multidisciplinary optimization software Isight.
[0030] In the aforementioned intelligent optimization method for the separation spring parameters of stacked satellites, the objective function for the multi-satellite separation safety assessment and the constraints of the optimization parameters in step three are used as the target, and the stiffness parameterization model of the separation spring in step two is substituted into iterative optimization.
[0031] In the above-mentioned intelligent optimization method for the separation spring parameters of stacked satellites, the method for selecting the separation spring optimization algorithm is as follows:
[0032] The multi-star separated spring constraint problem is a nonlinear constrained multi-parameter optimization problem. The generalized reduction gradient algorithm is selected as the separation spring optimization algorithm.
[0033] In the above-mentioned intelligent optimization method for the separation spring parameters of stacked satellites, in step five, after the final optimization of the inter-satellite separation dynamics simulation model between stacked satellites is completed, the optimization results are verified.
[0034] In the aforementioned intelligent optimization method for the separation spring parameters of stacked satellites, the separation safety is verified a second time after the spring parameters are rounded down in an engineering manner.
[0035] The advantages of this invention compared to the prior art are:
[0036] (1) This invention inherits the multibody dynamics simulation analysis software Adams and the multidisciplinary optimization software Adams through FMI. After building the multi-star separation simulation model, the multi-star separation spring parameter optimization model can be quickly established.
[0037] (2) The present invention uses a triaxial force method to simulate the mechanical characteristics of the separation spring. The advantage of this method compared with using spring units to simulate the mechanical characteristics of the spring is that it does not require the activation and disappearance of the spring force through sensors, which reduces the difficulty of writing scripts for multi-star separation simulation models.
[0038] (3) The advantage of this invention over the trial-and-error method, the enumeration method, and the empirical formula method is that the generalized reduction gradient algorithm can be used to solve the inter-satellite separation spring parameters in reverse, without repeated iterations, thus improving the efficiency of spring parameter selection. Attached Figure Description
[0039] Figure 1This is a flowchart illustrating the intelligent optimization process for the separation spring parameters of stacked satellites in this invention. Detailed Implementation
[0040] The present invention will be further described below with reference to the embodiments.
[0041] This invention provides an intelligent optimization method for the separation spring parameters of stacked satellites. It uses an intelligent optimization algorithm to solve the problem of selecting inter-satellite separation springs for stacked satellites. This method has high generalization and can be adapted to the selection and calculation of vertical separation spring parameters for various multi-satellite stacking systems.
[0042] Intelligent optimization method for multi-satellite separation spring parameters of stacked satellites, such as Figure 1 As shown, the specific steps include the following:
[0043] Step 1: Use multibody dynamics simulation analysis software to establish a dynamics simulation model of inter-satellite separation between stacked satellites.
[0044] Multibody dynamics software reads satellite 3D model entities in intermediate format from 3D modeling software to establish a dynamic simulation model of inter-satellite separation between stacked satellites.
[0045] Step 2: Establish a parameterized model of the stiffness of the separation spring in the separation dynamics simulation model, and use the parameterized model of the stiffness of the separation spring as the design variable.
[0046] The method for establishing the parameterized model of the stiffness of the release spring is as follows:
[0047] The mechanical properties of the release spring are simulated using a triaxial force approach, and the equation is as follows:
[0048]
[0049] In the formula, X i The spatial position of the free end of the separation direction spring before the separation of the i-th layer satellite;
[0050] X i+1 The coordinates of the point of application of the separation spring thrust before the separation of satellites in the i-th and (i+1)-th layers; X before multi-satellite separation i =X i+1 ;
[0051] L is the maximum length that the free end of the release spring can extend;
[0052] F i The separation spring pressure between the i-th layer satellite and the (i+1)-th layer satellite;
[0053] K i Let the spring stiffness be the separation stiffness of the i-th layer satellite;
[0054] L0 is the compression length of the release spring.
[0055] Step 3: In the parameterized model of the separation spring stiffness, establish the objective function for multi-star separation safety assessment and the optimization parameter constraints.
[0056] The objective function for the safety assessment of multi-satellite separation is:
[0057]
[0058] In the formula, Obj is the target value;
[0059] n is the total number of stacked satellite layers that are docked with the launch vehicle separation mechanism;
[0060] V i Let be the relative velocity between the i-th layer satellite and the launch vehicle separation mechanism;
[0061] The objective function for assessing the safety of multi-satellite separation is the sum of squares of the difference in separation velocities between two adjacent satellite layers minus 100 mm / s.
[0062] The optimization parameter constraints are as follows:
[0063] Among them, the lowest layer of satellites, which is in direct contact with the launch vehicle separation mechanism, separates from the mechanism at a speed greater than 200 mm / s; and the separation spring stiffness of the i-th layer of satellites is greater than 0, i.e., K i >0.
[0064] Step 4: Based on the objective function and optimization parameter constraints of the multi-satellite separation safety assessment in Step 3, select the separation spring optimization algorithm and iteratively optimize the stiffness parameterization model of the separation spring in Step 2.
[0065] This invention utilizes the FMI interface to integrate the multidisciplinary optimization software Isight and the multibody dynamics simulation analysis software. It quickly transfers the separation spring optimization parameters and objective function established in the multibody dynamics software to the multidisciplinary optimization software Isight, and combines them with the optimization algorithm function package in the multidisciplinary optimization software Isight.
[0066] The selection method for the spring separation optimization algorithm is as follows:
[0067] The multi-star separated spring constraint problem is a nonlinear constrained multi-parameter optimization problem. The generalized reduction gradient algorithm is selected as the separation spring optimization algorithm.
[0068] The objective function and optimization parameter constraints of the multi-satellite separation safety assessment in step three are used as the target, and they are substituted into the stiffness parameterization model of the separation spring in step two for iterative optimization.
[0069] Step 5: Complete the final optimization of the inter-satellite separation dynamics simulation model among stacked satellites.
[0070] After completing the final optimization of the simulation model of inter-satellite separation dynamics between stacked satellites, the optimization results were verified.
[0071] After the spring parameters are rounded down to the nearest integer, the separation safety is verified a second time.
[0072] Example
[0073] Step (1) Use multibody dynamics simulation analysis software to read the x_t format three-dimensional model data exported from CREO and build a dynamics simulation model of inter-satellite separation between stacked satellites.
[0074] Step (2) involves parametrically modeling the stiffness of the separation spring; its characteristics are:
[0075] The mechanical properties of the release spring were simulated by establishing a triaxial force, and its expression is as follows:
[0076]
[0077] In Equation 1), X i K represents the spatial position of the free end of the separation direction spring before the separation of the i-th layer satellite. i For the i-th layer satellite and K i Let X be the stiffness of the satellite separation spring in the (i+1)th layer. i+1 Let X be the coordinates of the point of application of the separation spring thrust before the separation of satellites in the i-th layer and (i+1)-th layer. i =X i+1 L0 is the compression length of the separation spring, and L is the maximum length that the free end of the separation spring can extend. The advantage of this method compared to using spring units to simulate the mechanical properties of springs is that it does not require sensors to control the activation and disappearance of spring forces, thus reducing the difficulty of writing scripts for multi-star separation simulation models.
[0078] Step (3) Construction of objective function and optimization parameter constraints for multi-star separation safety assessment.
[0079] The constructed objective function and optimization parameter constraints for multi-star separation safety assessment are as follows:
[0080]
[0081] Equations 2) and 3) use a 10-layer stacked satellite as an example, with the first layer of satellites docked with the launch vehicle separation mechanism. V i Let V1 be the relative velocity between the i-th layer satellite and the launch vehicle. The objective function of Equation 2) is the sum of squares of the difference in separation velocities between two adjacent layers of satellites minus 100 mm / s. V1 > 200 mm / s means that the separation velocity between the bottommost layer of satellites in direct contact with the launch vehicle and the launch vehicle is greater than 200 mm / s.
[0082] The separation spring parameter optimization model established in step (4) is to export the solution script files .acf and .cmd in Adams software, and to create a batch solution file .Bat based on the script language. Isight software establishes a data transmission channel between the simulation model and the optimization algorithm by reading and writing parameters. This method greatly reduces the cost of separation spring optimization modeling.
[0083] The optimization algorithm selected in step (5) is the generalized reduction gradient algorithm, which is one of the most effective methods for solving nonlinear optimization problems.
[0084] The generalized reduced gradient method simplifies the optimization problem to the following form:
[0085] Min f(K i ) 4)
[0086] Sth j (K i ) = 0, j = 1, 2, ..., m 5)
[0087] L≤K i ≤U 6)
[0088] The gradient calculation method of the simplified gradient algorithm differs from that of the traditional gradient algorithm, which uses the objective function f(K). i The partial derivatives with respect to the optimization variables are obtained by taking the partial derivatives with respect to the N-dimensional optimization variable K. i Divide into basis vectors and non-basis vectors.
[0089]
[0090] In the formula: B is an M-dimensional basis vector; N is a (NM) non-basis vector, and correspondingly, K i The upper limit U and lower limit L of the value can also be decomposed into two parts:
[0091] L = [L B L N ] T U = [U B U N ] T 8)
[0092] By the implicit function existence theorem, there exists a continuous mapping:
[0093]
[0094] Objective function f(K) i Transformed into:
[0095]
[0096] constraint function St
[0097]
[0098] The formula for calculating the generalized reduced gradient is as follows:
[0099]
[0100] Due to equality constraints:
[0101] For nonlinear constraints, Direct solution is difficult, so the elimination method is used. For constraint function h j (K i Taking the partial derivatives, we can obtain:
[0102]
[0103] Where p = 1, 2, ..., nm, j = 1, 2, ..., m.
[0104] Written in matrix form:
[0105]
[0106] The above equation is obtained by transformation. Substituting into the generalized reduced gradient calculation, we get
[0107]
[0108] Define the search direction S W W represents the number of optimization iterations.
[0109] when and or and At that time, S W =0;
[0110] Other situations:
[0111]
[0112] (a) Given allowable errors ε1, ε2 > 0, let W = 0; select the initial point K of the optimization variable. i 0 and divide it into
[0113] (b) Solving for the reduced gradient Solve for the search direction S W If ||S W If ||<ε1, then the objective function is considered to be... If the optimal solution has been reached, stop the optimization iteration; otherwise, proceed to the next step.
[0114] (c) Select Where α is any number greater than 0. If In the interval [L] N U N If the result is within the range indicated by α, proceed to the next step; otherwise, halve α and recalculate. until Falling in the interval [L N U N ]Inside.
[0115] (d) Solve the system of equations using Newton's method
[0116] make C = 1, D = 10;
[0117] (d1)
[0118] like and L B ≤y C+1 ≤U B Transfer to (e), or transfer to (4.2).
[0119] (d2) If C = D, then α is halved. C = 1; otherwise C = C + 1, return to (d1).
[0120] (e)x 0 =(y C+1 x N ), turn (a).
[0121] Step (6) After the spring parameters are engineered and rounded, a second verification of separation safety is performed.
[0122] This invention inherits the multibody dynamics simulation analysis software Adams and the multidisciplinary optimization software Adams through FMI. After building the multi-star separation simulation model, it can quickly establish a multi-star separation spring parameter optimization model.
[0123] This invention uses a triaxial force approach to simulate the mechanical properties of a separation spring. Compared to using spring units to simulate the mechanical properties of a spring, this method has the advantage of eliminating the need for sensors to control the activation and deactivation of the spring force, thus reducing the difficulty of writing scripts for multi-satellite separation simulation models.
[0124] The advantage of this invention over trial-and-error, enumeration, and empirical formula methods is that it can solve for the inter-satellite separation spring parameters by using the generalized reduction gradient algorithm, without repeated iterations, thus improving the efficiency of spring parameter selection.
[0125] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.
Claims
1. A method for intelligent optimization of spring parameters for multi-satellite separation in stacked satellites, characterized in that: include: Step 1: Establish a dynamic simulation model of inter-satellite separation between stacked satellites using multibody dynamics simulation analysis software; Step 2: Establish a parameterized model of the stiffness of the separation spring in the separation dynamics simulation model, and use the parameterized model of the stiffness of the separation spring as the design variable; Step 3: In the parameterized model of the separation spring stiffness, establish the objective function for multi-star separation safety assessment and the optimization parameter constraints; Step 4: Based on the objective function and optimization parameter constraints of the multi-satellite separation safety assessment in Step 3, select the separation spring optimization algorithm and iteratively optimize the stiffness parameterization model of the separation spring in Step 2. Step 5: Complete the final optimization of the inter-satellite separation dynamics simulation model among stacked satellites.
2. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 1, characterized in that: In step one, the multibody dynamics software reads the satellite 3D model entity in intermediate format from the 3D modeling software to establish a dynamic simulation model of inter-satellite separation between stacked satellites.
3. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 1, characterized in that: In step two, the method for establishing the stiffness parameterization model of the separation spring is as follows: The mechanical properties of the release spring are simulated using a triaxial force approach, and the equation is as follows: In the formula, X i This represents the spatial position of the free end of the separation direction spring before the separation of the i-th layer satellite; X i+1 The coordinates of the point of application of the separation spring thrust before the separation of satellites in the i-th and (i+1)-th layers; X before multi-satellite separation i =X i+1 ; L is the maximum length that the free end of the release spring can extend; F i The separation spring pressure between the i-th layer satellite and the (i+1)-th layer satellite; K i Let the spring stiffness be the separation stiffness of the i-th layer satellite; L0 is the compression length of the release spring.
4. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 3, characterized in that: In step three, the objective function for assessing the safety of multi-satellite separation is: In the formula, Obj is the target value; n is the total number of stacked satellite layers that are docked with the launch vehicle separation mechanism; V i Let be the relative velocity between the i-th layer satellite and the launch vehicle separation mechanism; The objective function for assessing the safety of multi-satellite separation is the sum of squares of the difference in separation velocities between two adjacent satellite layers minus 100 mm / s.
5. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 4, characterized in that: The constraints on the optimization parameters are as follows: Among them, the lowest layer of satellites, which is in direct contact with the launch vehicle separation mechanism, separates from the mechanism at a speed greater than 200 mm / s; and the separation spring stiffness of the i-th layer of satellites is greater than 0, i.e., K i >0.
6. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 5, characterized in that: In step four, the multidisciplinary optimization software Isight and the multibody dynamics simulation analysis software are integrated using the FMI interface. The separation spring optimization parameters and objective function established in the multibody dynamics software are quickly transferred to the multidisciplinary optimization software Isight, and combined with the optimization algorithm function package in the multidisciplinary optimization software Isight.
7. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 6, characterized in that: The objective function and optimization parameter constraints of the multi-satellite separation safety assessment in step three are used as the target, and they are substituted into the stiffness parameterization model of the separation spring in step two for iterative optimization.
8. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 7, characterized in that: The method for selecting the separation spring optimization algorithm is as follows: The multi-star separated spring constraint problem is a nonlinear constrained multi-parameter optimization problem. The generalized reduction gradient algorithm is selected as the separation spring optimization algorithm.
9. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 1, characterized in that: In step five, after the final optimization of the simulation model of inter-satellite separation dynamics between stacked satellites is completed, the optimization results are verified.
10. The intelligent optimization method for multi-satellite separation spring parameters of stacked satellites according to claim 9, characterized in that: After the spring parameters are rounded down to the nearest integer, the separation safety is verified a second time.