Highly scalable componentized track predictor and applications

Abstract

The application provides a highly scalable componentized orbit predictor and application, comprising: a model manager managing all dynamic submodels and interacting with an integrator, while supporting dynamic addition of new dynamic submodels through an external plug-in mechanism; in a prediction initialization stage, the model manager is responsible for establishing data interaction channels between dynamic submodels, collecting derivatives output by all dynamic submodels to count integral dimensions, and establishing a dynamic ordinary differential numerical model for the integrator to perform orbit prediction; an event detector is embedded in an integral progression process of the integrator and is bound to a corresponding action executor, the event detector judges whether an event is triggered by calculating a switch function, and executes the bound action executor when the event is triggered. The application solves the problem of scalability of the predictor, and can solve the problem of the predictor in different scenarios such as near-earth, earth-moon, deep space, etc. through flexible combination of different components.

Description

Technical Field

[0001] This invention relates to the field of aerospace mission design technology, specifically to a highly scalable modular orbit predictor and its applications. Background Technology

[0002] Orbit prediction is a fundamental function of space mission design software. Essentially, it predicts the future trajectory of a spacecraft under various dynamic influencing factors, generating time-series data such as position-velocity covariance for the software platform. Traditional orbit predictors typically hard-code dynamic models (such as Earth's gravity, solar radiation pressure, and atmospheric drag) into their code. When a new force model (such as a new upper atmosphere model) needs to be added, the entire software must be modified and recompiled, a cumbersome process prone to errors. To adapt to the versatility of mission design software, orbit predictors need to be compatible with different dynamic models and support rapid expansion of new functions to meet new requirements through script plugins and other means, without software updates. Summary of the Invention

[0003] To address the problems existing in the prior art, this invention provides a highly scalable modular orbit predictor and its application.

[0004] To achieve the above objectives, the present invention provides the following technical solution: On the one hand, a highly scalable modular orbit predictor is provided, including an integrator, an event detector, an action executor, a dynamic sub-model, a model manager, and an ephemeris collector; The model manager is used to manage all dynamic sub-models and interact with the integrator. It also supports the dynamic addition of new dynamic sub-models through an external plug-in mechanism. During the forecast initialization phase, it is responsible for establishing data interaction channels between dynamic sub-models and collecting the derivatives output by all dynamic sub-models to statistically measure the integration dimension, and establishing a dynamic ordinary differential numerical model for the integrator to perform orbit prediction. The event detector is embedded in the integral progression process of the integrator and is bound to at least one of the action executors. It is used to determine whether an event is triggered by calculating a switch function and to execute the bound action executor when the event is triggered. The action actuator is used to perform specified actions on objects in the aerospace mission design software; The ephemeris collector is used to convert continuous array data into orbital parameters and collect them according to the right function dimension of the integrator.

[0005] On the other hand, the application of the aforementioned highly scalable modular orbit predictor in orbit prediction tasks includes the following steps: Configure the core components of the framework, including: configuring the integrator as a fixed-step Runge-Kutta numerical integrator; configuring the model manager to load the basic geocentric gravity model; configuring the event detector as an endpoint time detector; configuring the action executor as a stop prediction executor; and configuring the ephemeris collector as a memory array collector. Set the forecast parameters and input the initial conditions for the orbit forecast through the configuration interface, including the initial position vector, initial velocity vector, forecast duration and integration step size, as well as the parameters of the dynamic sub-model; The orbit prediction is performed. The integrator calls the dynamic sub-model constructed by the model manager to perform numerical integration. During the integration process, the event detector monitors the prediction time. When the end time is reached, the action actuator is triggered to stop the calculation, and the ephemeris collector collects the orbit data. The orbit prediction task is completed by outputting the predicted orbital position and velocity vector sequence through the ephemeris collector.

[0006] On the other hand, the application of the aforementioned highly scalable modular orbit predictor in long-term orbit maintenance missions, through the collaboration of event detectors and actuators, achieves autonomous orbit control, including the following steps: Step 1, Configure the maintenance simulation component: Configure a dedicated component for long-term track maintenance tasks. Set up a track semi-major axis detector in the event detector and configure its trigger condition to be that the track semi-major axis decays to less than the preset maintenance threshold due to perturbation. Configure a track maneuver planner in the motion actuator, with a built-in track maintenance control algorithm and set the corresponding control parameters. Step 2, perform high-precision orbit prediction: The integrator calls the high-precision dynamic model built by the model manager to perform numerical integration; Step 3, Event Detection and Triggering: During the numerical integration process of the integrator, the event detector continuously monitors the evolution of the semi-major axis of the track. When the event detector detects the decay of the semi-major axis of the track and meets the preset triggering conditions, it immediately triggers the bound track maneuvering planner and transmits the current track state to the track maneuvering planner. Step 4, Execute orbit maintenance actions: After the orbit maneuver planner is triggered, it executes the following sub-steps: The orbit maneuver planner calculates the magnitude, direction, and timing of the required velocity increment based on the deviation between the current orbit state and the target orbit state; it applies the calculated velocity increment as an instantaneous pulse to the current spacecraft state vector; it updates the spacecraft mass parameters according to the thruster consumption model and generates a new spacecraft state vector after the orbit maneuver. Step 5: Continue the simulation process. The integrator uses the updated spacecraft state vector to continue integrating the prediction. Return to step 3 to achieve closed-loop control of "detection-planning-execution" until the total prediction time is reached.

[0007] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention designs the orbit predictor as a combination of the following components: integrator, event detector, action executor, dynamic sub-model, model manager, and ephemeris collector. This solves the scalability problem of the predictor and can solve predictor problems in different scenarios such as near-Earth, Earth-Moon, and deep space through flexible combination of different components. Furthermore, it can support new dynamic models by dynamically adding dynamic sub-models to the model manager.

[0008] Specifically, this invention modularizes dynamic sub-models and proposes a model manager that manages the dynamic sub-models through a data port registration and data binding mechanism. In this way, the dynamic model becomes an independent unit, and users only need to develop new dynamic sub-models according to the interface specifications to dynamically embed them through a plug-in approach.

[0009] This invention introduces a binding mechanism between event detectors and action executors, enabling a high degree of customizability and automation of the forecasting process.

[0010] This invention designs the entire orbit predictor as a combination of the following six components: integrator, event detector, action executor, dynamics sub-model, model manager, and ephemeris collector. These components communicate through well-defined interfaces, allowing each component to be developed, tested, replaced, and upgraded independently, greatly improving reusability and system maintainability. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a framework diagram of a highly scalable modular orbit predictor in one embodiment. Detailed Implementation

[0013] The technical solution of the present invention will now be clearly and completely described through specific embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0014] Reference Figure 1One embodiment provides a highly scalable modular orbit predictor, including an integrator, an event detector, an action executor, a dynamic sub-model, a model manager, and an ephemeris collector; The model manager is used to manage all dynamic sub-models and interact with the integrator. It also supports the dynamic addition of new dynamic sub-models through an external plug-in mechanism. During the forecast initialization phase, it is responsible for establishing data interaction channels between dynamic sub-models and collecting the derivatives output by all dynamic sub-models to statistically measure the integration dimension, and establishing a dynamic ordinary differential numerical model for the integrator to perform orbit prediction. The event detector is embedded in the integral progression process of the integrator and is bound to at least one of the action executors. It determines whether an event is triggered by calculating a switching function and executes the bound action executor when the event is triggered. The action actuator is used to perform specified actions on objects in the aerospace mission design software; The ephemeris collector is used to convert continuous array data into orbital parameters and collect them according to the right function dimension of the integrator.

[0015] In the above embodiments, the integrator is used to perform numerical integration or analytical calculations for orbit prediction, implement time stepping, and complete orbit prediction. An event detector is embedded in the integrator's solution process and bound to at least one of the action actuators. It determines whether an event meets a preset triggering condition by calculating a switching function; if so, it triggers the event and executes the corresponding bound action actuator. The action actuator is bound to the event detector and is used to perform a specified action when the event is triggered. A dynamic sub-model is used to provide various models in dynamic simulation, outputting model data and its derivatives based on different inputs. The model manager manages all dynamic sub-models, establishes data interaction channels between dynamic sub-models, and interacts with the integrator.

[0016] This invention proposes a data port registration and data binding mechanism for a dynamic sub-model. The model manager manages the dynamic sub-model through this mechanism, including: The dynamic sub-model provides a data port query interface, which is used to output the name, dimension, and connection port information of the input data, output data, and output derivative data of the dynamic sub-model. The model manager traverses all dynamic sub-models, collects all data information through the data port query interface of the dynamic sub-models, analyzes the dependency management between dynamic sub-models, establishes the connection between the data ports of dynamic sub-models, counts the number of non-integral parameters and integral parameters of the dynamic sub-models, and constructs the ordinary differential numerical model.

[0017] In one embodiment, the integrator iteratively calculates the time point at which the switching function of the event detector is 0, and after reaching a preset time convergence error, triggers the execution of the action executor bound to the event detector.

[0018] In any of the above embodiments, the event detector includes at least one of the following: duration detector, epoch detector, longitude detector, latitude detector, altitude detector, near apex detector, far apex detector, true near angle detector, and flat near angle detector.

[0019] In any of the above embodiments, the actions performed by the motion actuator include performing trajectory prediction, stopping trajectory prediction, performing pre-configured maneuvers, and updating model configuration parameters.

[0020] In any of the above embodiments, the ephemeris collector includes at least one of the following data output formats: hard disk file, memory array, dynamic two-dimensional chart, and dynamic data report.

[0021] Each component of the orbit predictor provided in the above embodiments supports customization and extension via plug-in scripts. Users can embed custom dynamic sub-models into the orbit predictor via plug-in scripts, including the following steps: Steps to prepare a plug-in model: Develop a plug-in model that conforms to the specifications, in which the input and output data are clearly defined; The steps for registering a plugin model are as follows: Add the plugin model to the model manager through the plugin registration interface. The model manager then calls the plugin model's data port query interface to obtain its data port information. Integration of plug-in models: The model manager analyzes the data dependencies between the plug-in model and the existing dynamic sub-model, establishes data connections between the input port of the plug-in model and the output port of the corresponding existing dynamic sub-model, summarizes the output of the plug-in model into the total output of the model manager, updates the derivative dimension of the dynamic sub-model, and notifies the integrator that the dynamic sub-model has been updated.

[0022] Furthermore, a verification and extension effect step can be added: execute an orbit prediction task, and verify that the plug-in model has been successfully integrated and is effective by using orbit data collected by the ephemeris collector.

[0023] Another embodiment provides the application of the highly scalable modular orbit predictor of the above embodiments in orbit prediction tasks, including the following steps: The core components of the configuration framework are initialized according to the mission requirements. Specifically, this includes: configuring the integrator as a fixed-step Runge-Kutta numerical integrator; configuring the model manager to load the basic Earth-centric gravity model; configuring the event detector as an endpoint time detector; configuring the action executor as a stop prediction executor; and configuring the ephemeris collector as a memory array collector.

[0024] The forecast parameters are set by inputting the initial conditions for orbit forecasting through the configuration interface, including the initial position vector, initial velocity vector, forecast duration, integration step size, and dynamic sub-model parameters. The dynamic sub-model parameters include the order of the Earth's gravity model, and configuration parameters for whether to enable three-body gravity, solar radiation pressure, and atmospheric drag. The model manager initializes the corresponding dynamic sub-model based on these parameters.

[0025] The orbit prediction is performed. The integrator calls the dynamic sub-model constructed by the model manager to perform numerical integration. During the integration process, the event detector monitors the prediction time. When the end time is reached, the action actuator is triggered to stop the calculation. The ephemeris collector collects the orbit data and stores it in the memory array.

[0026] The orbit prediction task is completed by outputting the predicted orbital position and velocity vector sequence through the ephemeris collector.

[0027] The highly scalable modular orbit predictor provided by this invention supports external plug-ins providing dynamic models. The following details how the highly scalable modular orbit predictor dynamically adds new dynamic sub-models through the external plug-in mechanism, illustrating the model's dynamic scalability. Specifically, dynamically adding new dynamic sub-models through the external plug-in mechanism includes the following steps: Step 1, Prepare the plug-in model: Develop a plug-in model that conforms to the interface specification to ensure interface compatibility. Plug-in models include, for example, atmospheric drag models and complex deep space dynamics models. This plug-in model clearly defines its input data (such as spacecraft ballistic coefficients, position vectors, velocity vectors, and atmospheric density) and output data (acceleration vectors and their corresponding coordinate systems).

[0028] Step 2, Register Plugin Models: Add plugin models (such as atmospheric drag models) to the model manager through the plugin registration interface. The model manager calls the plugin model's data port query interface to obtain its data port information, including data type, dimensions, and other parameters.

[0029] Step 3, Integrating the plug-in model: Taking the atmospheric drag model as an example, the model manager performs the following sub-steps to integrate the plug-in model: First, analyze the data dependency between the atmospheric drag model and the existing Earth gravity model to determine the data flow direction; then, establish the connection between the position vector output of the Earth gravity model and the input port of the atmospheric drag model; summarize the drag acceleration output of the atmospheric drag model into the total acceleration output of the model manager; update the derivative dimension of the dynamic sub-model and notify the integrator that the dynamic sub-model has been updated.

[0030] Step 4: Verify the extension effect and confirm functional integrity. Perform the same orbit prediction task to verify that the integrator automatically incorporates atmospheric drag effects during integration. Collect the extended predicted trajectories using the ephemeris collector. By comparing the orbit prediction results of the dynamic sub-model before and after extension, confirm the successful dynamic model extension. Thus, dynamic model extension was successfully achieved without modifying the core code, meeting the requirements for high-precision orbit prediction.

[0031] In one embodiment, the highly scalable modular orbit predictor described above is proposed for application in long-term orbit maintenance missions. Autonomous orbit control is achieved through the collaboration of event detectors and action actuators. Specifically, the method includes the following steps: Step 1, Configure the maintenance simulation component: Configure a dedicated component for long-term track maintenance tasks. Set up a track semi-major axis detector in the event detector and configure its trigger condition to be that the track semi-major axis decays to less than the preset maintenance threshold due to perturbation. Configure a track maneuver planner in the motion actuator, with built-in track maintenance control algorithms (such as the Hohmann transfer algorithm and track maintenance maneuver planning algorithm), and set the corresponding control parameters.

[0032] Step 2, perform high-precision orbit prediction: The integrator calls the high-precision dynamic model (including Earth's non-spherical gravity, solar and lunar perturbations, etc.) built by the model manager to perform numerical integration.

[0033] Step 3: Event Detection and Triggering: During numerical integration by the integrator, the event detector continuously monitors the evolution of the orbit's semi-major axis. When the event detector detects a decay in the orbit's semi-major axis and meets the preset triggering conditions, it immediately triggers the bound orbital maneuvering planner and transmits the current orbital state to the orbital maneuvering planner.

[0034] Step 4, Execute orbit maintenance actions: After the orbit maneuver planner is triggered, it executes the following sub-steps: The orbit maneuver planner calculates the magnitude, direction, and timing of the required velocity increment based on the deviation between the current orbit state and the target orbit state; it applies the calculated velocity increment as an instantaneous pulse to the current spacecraft state vector; and it updates the spacecraft mass parameters according to the thruster consumption model to generate a new spacecraft state vector after the orbit maneuver.

[0035] Step 5: Continue the simulation process. The integrator uses the updated spacecraft state vector to continue integrating the prediction. Return to step 3 to achieve closed-loop control of "detection-planning-execution" until the total prediction time is reached.

[0036] The highly scalable modular orbit predictor provided by this invention enables flexible data output and multi-format ephemeris collection. Specifically, the highly scalable modular orbit predictor diversifies data output through an ephemeris collector to meet the needs of different application scenarios. Specifically, diversifying data output through the ephemeris collector includes the following steps: Step 1, Configure multiple types of ephemeris collectors: Configure three or more ephemeris collectors simultaneously according to output requirements, including file output collectors (for data archiving), real-time visualization collectors (for situation display), and report generation collectors (for parameter analysis).

[0037] Step 2, Parallel Data Collection: During the orbit integration process performed by the integrator, multiple types of ephemeris collectors work in parallel, performing the following operations respectively: The file output collector writes the position, velocity vectors and timestamps to the hard disk file at preset time intervals; the real-time visualization collector maintains the dynamic three-dimensional trajectory in memory and refreshes the graphical interface in real time; the report generation collector calculates parameters such as orbital elements, energy and angular momentum in real time.

[0038] Step 3, Generate final output results: After the forecast mission is completed, each ephemeris collector generates the final output: File output collector closes the data file and completes data archiving; Real-time visualization collector freezes the final trajectory image; Report generation collector summarizes the orbital parameters of the entire forecast process and generates a comprehensive statistical report.

[0039] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made under the inventive concept of the present invention using the contents of the present invention specification and drawings, or direct / indirect applications in other related technical fields, are included within the protection scope of the present invention.

Claims

1. A highly scalable modular orbit predictor, characterized in that, Includes integrator, event detector, action executor, dynamics sub-model, model manager, and ephemeris collector; The model manager is used to manage all dynamic sub-models and interact with the integrator. It also supports the dynamic addition of new dynamic sub-models through an external plug-in mechanism. During the forecast initialization phase, it is responsible for establishing data interaction channels between dynamic sub-models and collecting the derivatives output by all dynamic sub-models to statistically measure the integration dimension, and establishing a dynamic ordinary differential numerical model for the integrator to perform orbit prediction. The event detector is embedded in the integral progression process of the integrator and is bound to at least one of the action executors. It is used to determine whether an event is triggered by calculating a switch function and to execute the bound action executor when the event is triggered. The action actuator is used to perform specified actions on objects in the aerospace mission design software; The ephemeris collector is used to convert continuous array data into orbital parameters and collect them according to the right function dimension of the integrator.

2. The highly scalable modular orbit predictor according to claim 1, characterized in that, The event detector includes at least one of the following: duration detector, epoch detector, longitude detector, latitude detector, altitude detector, near burrow detector, far burrow detector, true near angle detector, and flat near angle detector.

3. The highly scalable modular orbit predictor according to claim 1, characterized in that, The actions performed by the actuator include performing trajectory prediction, stopping trajectory prediction, performing pre-configured maneuvers, and updating model configuration parameters.

4. The highly scalable modular orbit predictor according to claim 1, characterized in that, The integrator iteratively calculates the time point at which the switching function of the event detector is 0. After reaching the preset time convergence error, it triggers the execution of the action executor bound to the event detector.

5. The highly scalable modular orbit predictor according to claim 1, characterized in that, The ephemeris collector includes at least one of the following data output formats: hard disk file, memory array, dynamic two-dimensional chart, and dynamic data report.

6. The highly scalable modular orbit predictor according to any one of claims 1 to 5, characterized in that, The model manager manages the dynamic sub-models through a data port registration and data binding mechanism, including: The dynamic sub-model provides a data port query interface for outputting the names, dimensions, and connection port information of the input data, output data, and output derivative data of the dynamic sub-model. The model manager traverses all dynamic sub-models, collects all data information through the data port query interface of the dynamic sub-models, analyzes the dependency management between dynamic sub-models, establishes the connection between the data ports of dynamic sub-models, counts the number of non-integral parameters and integral parameters of the dynamic sub-models, and constructs the ordinary differential numerical model.

7. The highly scalable modular orbit predictor according to claim 6, characterized in that, Users can embed custom dynamic sub-models into the orbit predictor via plug-in scripts.

8. The highly scalable modular orbit predictor according to claim 7, characterized in that, Embedding a custom dynamics sub-model into the orbit predictor includes the following steps: Steps to prepare a plug-in model: Develop a plug-in model that conforms to the specifications, in which the input and output data are clearly defined; The steps for registering a plugin model are as follows: Add the plugin model to the model manager through the plugin registration interface. The model manager then calls the plugin model's data port query interface to obtain its data port information. Integration of plug-in models: The model manager analyzes the data dependencies between the plug-in model and the existing dynamic sub-model, establishes data connections between the input port of the plug-in model and the output port of the corresponding existing dynamic sub-model, summarizes the output of the plug-in model into the total output of the model manager, updates the derivative dimension of the dynamic sub-model, and notifies the integrator that the dynamic sub-model has been updated.

9. The application of the highly scalable modular orbit predictor as described in claim 1 in orbit prediction tasks, characterized in that, Includes the following steps: Configure the core components of the framework, including: configuring the integrator as a fixed-step Runge-Kutta numerical integrator; configuring the model manager to load the basic geocentric gravity model; configuring the event detector as an endpoint time detector; configuring the action executor as a stop prediction executor; and configuring the ephemeris collector as a memory array collector. Set the forecast parameters and input the initial conditions for the orbit forecast through the configuration interface, including the initial position vector, initial velocity vector, forecast duration and integration step size, as well as the parameters of the dynamic sub-model; The orbit prediction is performed. The integrator calls the dynamic sub-model constructed by the model manager to perform numerical integration. During the integration process, the event detector monitors the prediction time. When the end time is reached, the action actuator is triggered to stop the calculation, and the ephemeris collector collects the orbit data. The orbit prediction task is completed by outputting the predicted orbital position and velocity vector sequence through the ephemeris collector.

10. The application of the highly scalable modular orbit predictor as described in claim 1 in long-term orbit maintenance missions, characterized in that, Autonomous orbit control is achieved through the collaboration of event detectors and motion actuators, including the following steps: Step 1, Configure the maintenance simulation component: Configure a dedicated component for long-term track maintenance tasks. Set up a track semi-major axis detector in the event detector and configure its trigger condition to be that the track semi-major axis decays to less than the preset maintenance threshold due to perturbation. Configure a track maneuver planner in the motion actuator, with a built-in track maintenance control algorithm and set the corresponding control parameters. Step 2, perform high-precision orbit prediction: The integrator calls the high-precision dynamic model built by the model manager to perform numerical integration; Step 3, Event Detection and Triggering: During the numerical integration process of the integrator, the event detector continuously monitors the evolution of the semi-major axis of the track. When the event detector detects the decay of the semi-major axis of the track and meets the preset triggering conditions, it immediately triggers the bound track maneuvering planner and transmits the current track state to the track maneuvering planner. Step 4, Execute orbit maintenance actions: After the orbit maneuver planner is triggered, it executes the following sub-steps: The orbit maneuver planner calculates the magnitude, direction, and timing of the required velocity increment based on the deviation between the current orbit state and the target orbit state; it applies the calculated velocity increment as an instantaneous pulse to the current spacecraft state vector; it updates the spacecraft mass parameters according to the thruster consumption model and generates a new spacecraft state vector after the orbit maneuver. Step 5: Continue the simulation process. The integrator uses the updated spacecraft state vector to continue integrating the prediction. Return to step 3 to achieve closed-loop control of "detection-planning-execution" until the total prediction time is reached.

Images