Satellite simulation method for multi-model joint variable step size solution

Abstract

The invention discloses a satellite simulation method for multi-model joint variable step size solution, which belongs to the technical field of satellite simulation, and is specifically implemented according to the following steps: S1, carrying out digital modeling and classification on each subsystem of a satellite; s2, inputting the classified model obtained in the S1 into a solver model, calculating self state parameters of each satellite entity at the moment t, and calculating interaction parameters between satellites by combining the self state parameters of each satellite entity; and S3, determining the time step n of the solver model, and returning to S2 to calculate the state parameters of the satellites and the interaction parameters between the satellites at the t + n moment. According to the method, the accuracy of all-digital satellite simulation calculation can be ensured, multi-solver parallel calculation and multi-satellite unified solution can be carried out, and the simulation efficiency is improved. According to the method, the causal relationship of the data is ensured through classified calculation of the satellite component model according to the causal relationship, dynamic request calculation of the step length and reinitialization of the integrator, and multi-model high-precision and high-efficiency mixed solution is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of satellite simulation, and in particular relates to a satellite simulation method for solving multiple models with joint variable step length. Background Art

[0002] At present, satellite simulation mainly adopts physical and semi-physical methods, which leads to large and complex simulation systems, low data exchange efficiency and slow simulation calculation speed. In recent years, the research on all-digital ultra-real-time satellite simulation system has gradually emerged. Satellite subsystems and components are all decomposed into different calculation processes or calculation modules, and multi-disciplinary calculation modules are synchronized with fixed step lengths. Shared memory, pipeline technology, message queues or sockets are used for data exchange. Due to the coupling relationship between models, the use of a longer fixed step length to exchange data will lose calculation accuracy. If a very small step length is used to exchange data, the calculation efficiency will be reduced, and it is difficult to achieve ultra-real-time simulation. At present, the module independent calculation simulation method based on the FMI simulation protocol and the distributed simulation method based on the HLA structure both have the problem of loss of calculation accuracy. The Chinese patent publication number is CN103926848B, and the patent name is: "Satellite simulation system and its data synchronization method", which adopts multi-module separate calculation and uses shared memory to synchronize data in the system discretization period. The above method will also lose calculation accuracy when there is a coupling relationship between the calculation modules. Summary of the invention

[0003] The purpose of the present invention is to provide a satellite simulation method for solving multiple models with variable step-size, which solves the technical problem that the existing simulation system is difficult to balance the calculation accuracy and calculation efficiency.

[0004] The technical solution adopted by the present invention is a satellite simulation method for solving multiple models with variable step sizes, which is specifically implemented according to the following steps:

[0005] S1, digital modeling and classification of satellite subsystems;

[0006] S2, input the classified model obtained in S1 into the solver model, calculate the state parameters of each satellite entity at time t, and calculate the interaction parameters between satellites in combination with the state parameters of each satellite entity;

[0007] S3, determine the solver model time step n, and return to S2 to calculate the satellite's own state parameters and the interaction parameters between satellites at time t+n.

[0008] The present invention is also characterized in that:

[0009] The satellite subsystems in S1 include satellite control subsystem, thermal control subsystem, power subsystem, payload subsystem, propulsion subsystem and structural subsystem.

[0010] In S1, each satellite subsystem is digitally modeled and classified, including: environment model, motion model, perception model, decision model and action model.

[0011] The solver model includes an internal solver model and an external solver model. The environment model and the motion model are input into the internal solver model, and the perception model, the decision model and the action model are input into the external solver model. The internal solver model takes precedence over the external solver model for solving.

[0012] The specific calculation of each satellite entity's own state parameters and the interaction parameters between satellites in S2 is:

[0013] Through the discrete equation, we can directly solve it by algebraic operation, and take any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. i , parameter v i The value of is calculated from the parameters and time associated with it, as shown in formula (1):

[0014] v i (t)=F(u1,u2,...u n ,t) (1)

[0015] In formula (1), u1,u2,...,u n Represents and parameter v i Related parameters;

[0016] Combined with the classified model obtained in S1, we get the parameter v i .

[0017] The specific calculation of each satellite entity's own state parameters and the interaction parameters between satellites in S2 is:

[0018] By performing integral operation on differential equations, we can solve the problem by taking any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. j , parameter v j Expressed as the derivative of its associated parameter, i.e. parameter v j It is obtained by integral calculation, as shown in formula (2):

[0019]

[0020] In formula (2), u1,u2,...,u m Indicates that v j Parameters related to parameters, combined with the classified model obtained in S1, get parameter v j .

[0021] The specific process of determining the solver model time step n in S3 is:

[0022] Set the system time step n s , action model time step n ai , the perception model requests time step n si , the minimum time step is used as the solver model time step n.

[0023] The beneficial effects of the present invention are:

[0024] The satellite simulation method of the present invention with multi-model joint variable step-size solution can ensure the accuracy of all-digital satellite simulation calculation, can perform multi-solver parallel calculation, and can solve multiple satellites uniformly, thereby improving simulation efficiency. The present invention ensures the causal relationship of data by classifying and calculating satellite component models according to causal relationships, dynamically requesting calculation steps, and reinitializing integrators, thereby achieving high-precision and high-efficiency hybrid solutions of multiple models. DETAILED DESCRIPTION

[0025] The present invention will be further explained below in conjunction with specific embodiments.

[0026] Example 1

[0027] The satellite simulation method for solving multiple models with variable step-size is specifically implemented according to the following steps:

[0028] S1, digital modeling and classification of satellite subsystems;

[0029] S2, input the classified model obtained in S1 into the solver model, calculate the state parameters of each satellite entity at time t, and calculate the interaction parameters between satellites in combination with the state parameters of each satellite entity;

[0030] S3, determine the solver model time step n, and return to S2 to calculate the satellite's own state parameters and the interaction parameters between satellites at time t+n.

[0031] In view of the problem that the current all-digital satellite simulation system cannot take into account both calculation accuracy and calculation efficiency, the present invention classifies the calculation models of various satellite components, adds each model to the internal solver and the external solver respectively according to the causal relationship of the model, solves the model in the internal solver at the same step point first, and then solves the model in the external solver, thereby ensuring the causal relationship of the data and realizing multi-model high-precision hybrid solution. Each satellite component dynamically requests the calculation step according to the time of instruction execution (such as orbit change instructions accurate to milliseconds) and event calculation (such as satellite collision events accurate to meters), ensuring the time accuracy of instructions and events. At the same time, during the simulation process, in view of the problem of inaccurate internal state of the integrator caused by the dynamic change of differential equations (such as a satellite exiting the simulation scene), a dynamic reinitialization method of the integrator is proposed to improve the solution accuracy. In the process of digital simulation of multiple satellites, the present invention uses the same solver to solve the state parameters of the satellite control subsystem, thermal control subsystem, power subsystem, load subsystem, propulsion subsystem, and structural subsystem. The method has higher calculation accuracy than the method of decomposition calculation and timed data exchange, and improves the calculation accuracy of all-digital satellite simulation.

[0032] Example 2

[0033] On the basis of Example 1, the satellite subsystems in S1 include a satellite control subsystem, a thermal control subsystem, a power subsystem, a payload subsystem, a propulsion subsystem and a structural subsystem.

[0034] Specifically, S1 digitally models and classifies the satellite control subsystem, thermal control subsystem, power subsystem, payload subsystem, propulsion subsystem and structural subsystem, including:

[0035] Environmental models, specifically the conventional celestial motion model in the simulation scene, as well as the atmospheric model, electromagnetic environment model, and solar magnetic storm model that affect the implementation of the simulation task;

[0036] Motion model, including the calculation model of the position, speed, posture, working status and other changes of the simulated entity in the simulation scene;

[0037] Perception model, direct perception parameters of the environment and other entities, such as collision detection, direct communication, etc., or energy-bearing media;

[0038] Decision-making models are used to simulate entities to make decisions based on their own state, the state of the environment and other entities, and perception data, and according to decision-making algorithms such as automatic control and artificial intelligence;

[0039] The action model is used to simulate the execution of the action instructions of the entity, change its own state, and generate new entities and energy.

[0040] In this embodiment, the above models all include: an interface for calculating the state at the initial moment; an interface for calculating all states at the current moment; an interface for outputting the current state; an interface for obtaining the current value of the integral state variable; an interface for setting the state vector at the step point given by the integrator to the model state variable; an interface for calculating all derived state parameters at the step point given by the integrator; and an interface for writing the derivative value of the model state vector into the derivative vector. The above interfaces are all connected to the solver model.

[0041] Example 3

[0042] Based on Example 2, the solver model includes an internal solver model and an external solver model. The environment model and the motion model are input into the internal solver model, and the perception model, the decision model and the action model are input into the external solver model. The internal solver model takes precedence over the external solver model for solving.

[0043] Due to the causal relationship between simulation data, some derived state parameters depend on the core state parameters. Therefore, the core state parameters are calculated using the internal solver first, and then the derived state parameters are calculated using the external solver at the same simulation time. This can ensure that some derived state parameters obtain theoretically accurate values.

[0044] Example 4

[0045] In the process of multi-satellite entity simulation, there are interactions between satellites, such as observation, communication, collision and other external interactions. The calculation of such interactions depends on the state parameters of each satellite entity, so it is necessary to ensure that the accurate current state is used to calculate the interaction relationship after the state of each satellite entity itself is calculated. Therefore, based on Example 3, the calculation of the state parameters of each satellite entity itself and the calculation of the interaction parameters between satellites in S2 are specifically:

[0046] Through the discrete equation, we can directly solve it by algebraic operation, and take any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. i , parameter v i The value of is calculated from the parameters and time associated with it, as shown in formula (1):

[0047] v i (t)=F(u1,u2,...u n ,t) (1)

[0048] In formula (1), u1,u2,...,u n Represents and parameter v i Related parameters;

[0049] Combined with the classified model obtained in S1, we get the parameter v i .

[0050] Example 5

[0051] On the basis of Example 3, different from Example 4, the calculation of each satellite entity's own state parameter and the calculation of the interaction parameters between satellites in S2 are specifically:

[0052] By performing integral operation on differential equations, we can solve the problem by taking any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. j , parameter v j Expressed as the derivative of its associated parameter, i.e. parameter v j It is obtained by integral calculation, as shown in formula (2):

[0053]

[0054] In formula (2), u1,u2,...,u m Indicates that v j Parameters related to parameters, combined with the classified model obtained in S1, get parameter v j .

[0055] Example 6

[0056] In the same integrator, all parameters of the differential equations to be solved for each solver model set are integrated and solved, and all parameters of the algebraic equations to be solved for the model set are algebraically solved. Specifically, based on Example 4 or Example 5, the specific process of determining the solver model time step n in S3 is:

[0057] Set the system time step n s , action model time step n ai , the perception model requests time step n si , the minimum time step is used as the solver model time step n.

[0058] Parameter coupling is common in satellite dynamics, thermal control, and electronic control subsystems. Combining multiple differential equations into the same integrator for joint integration will improve the computational accuracy of digital simulation.

[0059] The above is only a specific implementation of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A satellite simulation method using multiple models and variable step-size solutions, characterized in that: Follow the steps below to implement it: S1, digital modeling and classification of satellite subsystems; S2, input the classified model obtained in S1 into the solver model, calculate the state parameters of each satellite entity at time t, and calculate the interaction parameters between satellites in combination with the state parameters of each satellite entity; S3, determine the solver model time step n, and return to S2 to calculate the satellite's own state parameters and the interaction parameters between satellites at time t+n.

2. The satellite simulation method of claim 1, wherein: The satellite subsystems described in S1 include satellite control subsystem, thermal control subsystem, power subsystem, payload subsystem, propulsion subsystem and structural subsystem.

3. The satellite simulation method of multi-model joint variable step-size solution according to claim 2 is characterized in that: The digital modeling and classification of satellite subsystems described in S1 specifically include: environment model, motion model, perception model, decision model and action model.

4. The satellite simulation method of multi-model joint variable step-size solution according to claim 3 is characterized in that: The solver model includes an internal solver model and an external solver model. The environment model and the motion model are input into the internal solver model, and the perception model, the decision model and the action model are input into the external solver model. The internal solver model takes precedence over the external solver model in solving.

5. The satellite simulation method of claim 4 with multiple models and variable step-size solution, characterized in that: The calculation of the state parameters of each satellite entity itself and the calculation of the interaction parameters between satellites in S2 are specifically: Through the discrete equation, we can directly solve it by algebraic operation, and take any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. i , parameter v i The value of is calculated from the parameters and time associated with it, as shown in formula (1): v i (t)=F(u1,u2,...u n ,t) (1) In formula (1), u1,u2,...,u n Represents and parameter v i Related parameters; Combined with the classified model obtained in S1, we get the parameter v i .

6. The satellite simulation method of claim 4 with multiple models and variable step-size solution, characterized in that: The calculation of the state parameters of each satellite entity itself and the calculation of the interaction parameters between satellites in S2 are specifically: By performing integral operation on differential equations, we can solve the problem by taking any parameter v among the state parameters of each satellite entity and the interaction parameters between the satellites. j , parameter v j Expressed as the derivative of its associated parameter, i.e. parameter v j It is obtained by integral calculation, as shown in formula (2): In formula (2), u1,u2,...,u m Indicates that v j Parameters related to parameters, combined with the classified model obtained in S1, get parameter v j .

7. The satellite simulation method of multiple models combined with variable step-size solution according to claim 5 or 6, characterized in that: The specific process of determining the solver model time step n described in S3 is: Set the system time step n s , action model time step n ai , the perception model requests time step n si , the minimum time step is taken as the solver model time step n.

Images