A method based on mixed-dimensional space situation visualization system

Abstract

The present invention discloses a method based on a mixed-dimensional space situation visualization system, and relates to the technical field of space situation visualization. The present invention adopts dynamic dimensionality reduction technology and an adaptive weight allocation algorithm, and adjusts the dimensionality reduction strategy according to the dynamic changes of data in each dimension, so as to effectively capture complex nonlinear relationships, retain important characteristic dimensions, and avoid losing key situation information while reducing the dimensions. The digital twin technology is introduced to combine historical and real-time data, and an accurate satellite dynamic model is generated in real time under the conditions of orbit, attitude and external disturbance, and the orbit and attitude of the satellite are tracked in real time. The future trajectory and attitude are dynamically predicted based on the prediction model of orbital mechanics equations and external disturbances, covering short-term and long-term change trends, and the collision risk is dynamically calculated based on the relative position and relative speed model of the satellite and space debris, and the weight is adjusted in combination with the adaptive attenuation coefficient to update the collision risk assessment in real time.

Description

Technical Field

[0001] The present invention relates to the field of space situation visualization technology, and in particular to a method based on a mixed-dimensional space situation visualization system. Background Art

[0002] Space capabilities have become the strategic commanding heights in the space field. Space situational awareness is the process of recognizing and analyzing various factors that affect space activities. In a broad sense, space situational awareness covers the ability to perceive all events, threats, activities and states in space, ensuring that commanders, decision makers and operators can gain and maintain space advantages.

[0003] Space situation visualization involves processing and displaying multi-dimensional, large-scale, and dynamically changing space data. Multi-dimensionality is reflected in dimensions such as satellite orbit parameters, attitude control, speed, satellite health status, space debris, etc. Such data not only has many dimensions, but also has complex relationships with each other. Therefore, these data need to be compressed to facilitate effective display in the visualization interface. In addition, the orbits and states of space objects (satellites, space debris, spacecraft, etc.) are constantly changing and are affected by a variety of external factors (such as the earth's gravity field, solar radiation, and space weather). Such dynamic changes involve changes in position and speed, and are accompanied by attitude adjustments and changes in energy states. In collision avoidance and orbit planning tasks, tracking such dynamic evolution is crucial for space situation awareness.

[0004] Traditional display schemes usually use linear dimensionality reduction algorithms such as PCA to process high-dimensional data. PCA extracts the principal components in the data and projects the high-dimensional data into a lower-dimensional space. However, nonlinear relationships in the data, such as nonlinear perturbations of satellite orbits, will be simplified or lost during the dimensionality reduction process and cannot be displayed correctly. Some nonlinear dimensionality reduction algorithms such as tSNE (t-distributed neighborhood embedding) and UMAP (unified manifold approximation and projection) can capture local structures in complex data when dealing with nonlinear relationships. However, such algorithms have high computational complexity and are difficult to meet real-time requirements when processing large amounts of space situation data that are updated in real time. For tracking the dynamic evolution of classes, most common solutions are based on historical data. Trajectory prediction is performed based on static models, but the influence of various nonlinear interferences, such as solar wind and magnetic field effects, is ignored in long-term predictions. At the same time, traditional space situation visualization systems can only display the movement of objects through a timeline, but it is difficult to effectively track the multi-dimensional dynamic changes of objects, such as the attitude and health status of satellites. Some visualization solutions use prediction models based on historical data to track the trajectory changes of objects. With the timeline tool, the past and future trajectories of objects can be viewed within a certain range. However, this method is difficult to capture key situation information in dynamic changes when facing long-term nonlinear changes. Therefore, there is an urgent need for a visualization system based on mixed-dimensional space situation and its visualization method to solve such problems. Summary of the invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, the present invention provides a method based on a mixed-dimensional space situation visualization system to solve the problems that nonlinear information is often lost in the process of high-dimensional data compression and the real-time performance is insufficient; the dynamic evolution tracking is limited by the prediction model and cannot accurately reflect the complex orbit and attitude changes, making it difficult to fully grasp the dynamic situation of space objects, thus affecting the quality of decision-making.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] The present invention provides a method based on a mixed-dimensional space situation visualization system, comprising:

[0009] Step S1, space situation data collection, obtain real-time space situation data from sensors, input the space situation data into the twin simulation system, create a digital twin of each satellite, and synchronize the digital twin with the physical satellite status in real time

[0010] Step S2, multi-level dimensional compression, extracts important feature data in the space situation based on the digital twin, and dynamically reduces and compresses the multi-dimensional data;

[0011] Step S3, dynamic evolution tracking, dynamic evolution prediction based on the reduced-dimensional data, and integrating the results into visual display;

[0012] The system framework adopted by the above solution is:

[0013] The data acquisition module is responsible for acquiring real-time space situation data and inputting the data into the twin simulation system. The space situation data includes the satellite's orbit information, speed, attitude, near-Earth environment, and dynamic situation information of space debris;

[0014] The twin simulation system generates a corresponding digital twin based on the real-time data of each satellite, and the digital twin is synchronized with the physical satellite status in real time;

[0015] Dimensional compression module, which uses the twin simulation system to dynamically extract and compress multi-dimensional data in space situation;

[0016] Important feature identification module, used to highlight the relevant dimensions of important feature parameters;

[0017] Dynamic dimensionality reduction module, which dynamically adjusts the dimensionality reduction algorithm based on the prediction results of twin simulation;

[0018] Dynamic evolution tracking and simulation prediction module, which is based on the twin simulation system, performs real-time prediction, tracks the future dynamic changes of the trajectory and attitude of space objects, updates the object status in real time, and integrates the change information into the visualization interface;

[0019] Visual interface for visual display of multi-dimensional data;

[0020] Furthermore, in step S1, the steps of creating the satellite digital twin are:

[0021] Define track element status: Where r(t) represents the position vector of the satellite at time t, a represents the semi-major axis of the orbit, e represents the orbit eccentricity, and θ(t) represents the true anomaly of the satellite at time t.

[0022] Modeling satellite velocity: Where v(t) represents the velocity of the satellite at time t, μ represents the gravitational parameter of the Earth, r(t) represents the distance from the satellite to the center of the Earth, and a represents the semi-major axis of the orbit;

[0023] The attitude control model is built. The attitude of the satellite is represented by Euler angles, and the attitude change model is represented based on the rigid body dynamics equation. Where I represents the satellite's moment of inertia matrix, ω represents the satellite's angular velocity vector, and T represents the external torque on the satellite;

[0024] Modeling of the near-earth environmental impact: where represents the force caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj represents the projection area of ​​the satellite, which changes dynamically according to the attitude of the satellite n(t), c represents the speed of light, and n(t) represents the normal direction vector of the satellite surface;

[0025] Dynamic modeling of space debris using a distance-based risk model:

[0026] Among them, R collision (t) represents the collision risk at time t, represents the position vector of the satellite at time t, and r debris (t) represents the position vector of the space debris at time t, σ represents the standard deviation of the position uncertainty, which is estimated based on the sensor accuracy and error; the digital twin can track the dynamic state of the satellite in real time and can adjust its orbit and attitude under the influence of external disturbances.

[0027] Furthermore, in step S2, the twin simulation system extracts features from the dynamic data in the space situation, identifies and highlights important dynamic change dimensions based on historical and real-time data, including orbital deviation, attitude adjustment, and potential collision risk, extracts key information dimensions, and uses an adaptive weight allocation algorithm to score the importance of each dimension of data, and obtains multi-dimensional feature data containing dynamic weight allocation results;

[0028] Dynamically select dimensionality reduction algorithm based on multi-dimensional feature data with dynamic weight assignment results.

[0029] Furthermore, in step S2, the dynamic data feature extraction method in the space situation is:

[0030] Extract the characteristics of orbital deviation, Δr(t) = r(t)-r ref (t), where Δr(t) represents the deviation of the satellite from the reference orbital position at time t, r(t) represents the real-time position of the satellite at time t, and r ref (t) represents the reference orbital position based on historical data. The orbital deviation amplitude is in, represents the rate of change of the semi-major axis of the orbit, represents the rate of change of orbital eccentricity;

[0031] Feature extraction for posture adjustment, Δω(t)=ω(t)-ω ref (t), where Δω(t) represents the difference between the real-time angular velocity and the reference angular velocity, ω(t) represents the real-time angular velocity of the satellite at time t, and ω ref(t) represents the reference angular velocity based on historical data;

[0032] The Euler angle change of the posture is, Δφ(t)=φ(t)-φ ref (t),Δθ(t)=θ(t)-θ ref (t),Δψ(t)=ψ(t)-ψ ref (t), where Δφ(t), Δθ(t), and Δψ(t) represent the differences between the three Euler angles of the satellite attitude and the reference Euler angle, φ(t), θ(t), and ψ(t) represent the Euler angles of the satellite attitude at time t, and φ ref (t),θ ref (t),ψ ref (t) represents the reference Euler angle based on historical data.

[0033] Furthermore, in step S2, the method for extracting dynamic data features in the space situation also includes:

[0034] Extract the features of potential collision risk based on the relative distance and relative speed between the satellite and space debris. rel (t)=||r ref (t)-r debris (t)||, where d rel (t) represents the relative distance between the satellite and the space debris at time t, r ref (t) represents the position vector of the satellite at time t, r debris (t) represents the position vector of the space debris at time t, and the relative velocity is v rel (t)=||v sat (t)-v debris (t)||, where v rel (t) represents the relative speed between the satellite and the space debris, v sat (t) represents the velocity vector of the satellite at time t, v debris (t) represents the velocity vector of the space debris at time t;

[0035] Combined relative distance d rel (t) and relative speed v rel (t), calculate the probability of collision risk, Among them, R collision (t) represents the collision risk at time t. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, σ v Represents the velocity attenuation coefficient, calculate R collision (t), dynamically assess the collision risk between satellites and space debris.

[0036] Furthermore, in step S2, the method for extracting dynamic data features in the space situation also includes:

[0037] Use the adaptive weight assignment algorithm to score the importance of each dimension data, and define the importance weight of the dimension as w i (t), the importance of each dimension at time t is Among them, w i (t) represents the weight of the i-th dimension at time t, Δf i (t) represents the change in the i-th dimension, including orbital deviation, attitude adjustment or collision risk change, and N represents the total number of dynamic dimensions;

[0038] Weight w i (t) Ensure that important dynamically changing dimensions in multidimensional data are highlighted;

[0039] According to the dimension weight w i (t), dynamically adjusted nonlinear dimensionality reduction techniques tSNE and UMAP are selected;

[0040] The cost function is used in the dimensionality reduction process to retain the important information dimensions:

[0041] in, represents the loss function in the dimensionality reduction process, represents the eigenvector of the i-th dimension in high-dimensional space, Represents the feature vector of the i-th dimension in the low-dimensional space, and the loss function is based on the weight w i (t) Dynamic adjustment to effectively extract dynamic features in the space situation and highlight important dimensions (orbit, attitude, collision risk).

[0042] Furthermore, in step S3, a dynamic prediction of the future situation is performed based on the compressed data of step S2, the predicted data is integrated with the current situation information, the dynamic changes of the space object are tracked, and the predicted results of the future trajectory and situation are output to provide dynamic evolution data support for the visualization interface, which not only displays the current situation, but also performs dynamic prediction based on twin simulation to effectively track the orbit and attitude changes of the object.

[0043] Furthermore, the dynamic prediction method of future situation based on compressed data in step S3 is:

[0044] Using the compressed data in step S2, combined with the orbit, attitude and potential collision risk key feature dimensions, a dynamic prediction of the future space situation is made, the steps including orbit prediction, attitude prediction and collision risk prediction;

[0045] Orbital dynamics prediction: Where r(t+Δt) represents the position of the satellite at the future time t+Δt, r(t) represents the position of the satellite at the current time t, v(t) represents the velocity of the satellite at the current time t, a(t) represents the acceleration of the satellite at the time t, Δt represents the prediction time step, and the orbital position is predicted using the current velocity v(t) and acceleration a(t), where the acceleration a(t) is modeled by gravity and external disturbances, and the disturbance term is dynamically adjusted based on historical data and simulation models;

[0046] The acceleration of the gravitational disturbance is expressed as Among them, a gravity (t) represents the gravitational acceleration vector, μ represents the earth's gravitational constant, r(t) represents the satellite's current position vector, and the acceleration is calculated by the pull of the earth's gravity on the satellite, and changes dynamically with the position r(t);

[0047] Use rigid body dynamics equations and historical posture data to perform posture dynamics prediction:

[0048] θ(t+Δt)=θ(t)+ω(t)Δt, where θ(t+Δt) represents the attitude angle of the satellite at the future time t+Δt, θ(t) represents the attitude angle of the satellite at the current time t, ω(t) represents the angular velocity vector of the satellite at the current time t, and Δt represents the prediction time step.

[0049] Furthermore, the method of dynamically predicting future trends based on compressed data in step S3 also includes:

[0050] To model the attitude disturbance, Among them, T srp (t) represents the torque caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj (t) represents the projected area of ​​the satellite, which depends on the attitude angle θ(t), n(t) represents the normal direction vector of the satellite surface, which depends on the current attitude, c represents the speed of light, and the angular acceleration α(t) caused by the torque is: α(t) = I -1 T srp (t), where α(t) represents the attitude angular acceleration, I represents the satellite’s moment of inertia matrix, and T srp (t) represents the torque caused by the solar radiation pressure, which is numerically integrated and the angular velocity ω(t) is updated using the angular acceleration α(t), thereby updating the satellite's attitude angle θ(t);

[0051] Dynamically predict potential collision risks, based on the relative position and relative speed of the satellite and space debris, predict future collision risks, relative distance d rel (t) and relative velocity v rel The evolution of (t) is predicted to be: d rel(t+Δt)=d rel (t)+v rel (t)Δt, where d rel (t+Δt) represents the relative distance between the satellite and the debris at the future time t+Δt, d rel (t) represents the relative distance at the current time t, v rel (t) represents the relative speed at the current time t, and Δt represents the prediction time step;

[0052] Relative speed v at future time rel (t+Δt) is adjusted by the external disturbance prediction model, v rel (t+Δt)=v rel (t)+a rel (t)Δt, where v rel (t+Δt) represents the relative speed at the future time t+Δt, a rel represents relative acceleration;

[0053] Based on the dynamic changes in relative distance and speed, the future collision risk is calculated as follows:

[0054] Among them, R collision (t+Δt) represents the collision risk at the future time t+Δt. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, which controls the weight of relative distance on collision risk, σ v represents the speed attenuation coefficient, which controls the weight of relative speed on collision risk.

[0055] Integrate dynamic prediction with current situation,

[0056] S future (t+Δt)=λS current (t)+(1-λ)+S predicted (t+Δt), where S future (t+Δt) represents the future situation after fusion, combining current and predicted data, S current (t) represents the current situation data (orbit, attitude, collision risk), S predicted (t+Δt) represents the predicted situation data at the future moment, and λ represents the smoothing coefficient, which controls the fusion ratio between the current situation and the future prediction. The current situation is combined with the prediction results to ensure the accuracy of the prediction and the consistency of real-time updates.

[0057] The beneficial effects of the present invention are:

[0058] The present invention adopts dynamic dimensionality reduction technology and adaptive weight allocation algorithm, and adjusts the dimensionality reduction strategy according to the dynamic changes of data in each dimension. It can effectively capture complex nonlinear relationships, retain important feature dimensions, and avoid losing key situation information while reducing dimensions.

[0059] The present invention introduces digital twin technology to combine historical and real-time data, generates an accurate satellite dynamic model in real time under the conditions of orbit, attitude and external disturbance, tracks the satellite's orbit and attitude in real time, and dynamically predicts the evolution of future trajectory and attitude based on orbital mechanics equations and external disturbance prediction models, covering short-term and long-term change trends.

[0060] The present invention dynamically calculates the collision risk based on the relative position and relative speed model of the satellite and space debris, and adjusts the weight in combination with the adaptive attenuation coefficient to update the collision risk assessment in real time. It can issue an early warning during the rapid calculation and adjustment process to indicate the potential collision risk, especially in the case of multi-dimensional environmental changes, to ensure the safe operation of the satellite.

[0061] The present invention introduces modeling based on solar radiation pressure and external disturbance of the geomagnetic field to update the orbit and attitude of the satellite in real time. The adaptive adjustment mechanism enables the system to maintain high flexibility and robustness when facing a complex external environment, and can remain stable and adjust the orbit and attitude parameters in time even when subjected to external disturbances.

[0062] The present invention dynamically compresses and reduces the dimensionality of multi-dimensional feature data, compressing complex trajectory, posture and risk dimensions into easy-to-understand visual information.

[0063] The present invention utilizes the synchronization mechanism of digital twins and physical satellite states, as well as the dynamic fusion of historical and real-time data, which greatly improves the real-time performance and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0065] Figure 1 It is a schematic diagram of the structure of the space situation visualization system based on mixed dimensions of the present invention;

[0066] Figure 2 It is a flow chart of the visualization method based on the mixed-dimensional space situation system of the present invention. DETAILED DESCRIPTION

[0067] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0068] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0069] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0070] Example 1, reference Figure 1 , this embodiment provides a space situation visualization system based on mixed dimensions, including:

[0071] The data acquisition module is responsible for acquiring real-time space situation data and inputting the data into the twin simulation system. The space situation data includes the satellite's orbit information, speed, attitude, near-Earth environment, and dynamic situation information of space debris;

[0072] The twin simulation system generates a corresponding digital twin based on the real-time data of each satellite, and the digital twin is synchronized with the physical satellite status in real time;

[0073] Dimensional compression module, which uses the twin simulation system to dynamically extract and compress multi-dimensional data in space situation;

[0074] Important feature identification module, used to highlight the relevant dimensions of important feature parameters;

[0075] Dynamic dimensionality reduction module, which dynamically adjusts the dimensionality reduction algorithm based on the prediction results of twin simulation;

[0076] Dynamic evolution tracking and simulation prediction module, which is based on the twin simulation system, performs real-time prediction, tracks the future dynamic changes of the trajectory and attitude of space objects, updates the object status in real time, and integrates the change information into the visualization interface;

[0077] Visual interface for visual display of multi-dimensional data.

[0078] Example 2, reference Figure 2 , this embodiment provides a method based on a mixed-dimensional space situation visualization system, comprising the following steps:

[0079] Step S1, space situation data collection, obtain real-time space situation data from sensors, input the space situation data into the twin simulation system, create a digital twin of each satellite, and synchronize the digital twin with the physical satellite status in real time

[0080] Specifically, a digital twin is created to monitor the satellite’s orbit, attitude and changes in the external environment in real time. The digital twin is updated in real time based on the orbit, speed, acceleration and external disturbances, synchronized with the physical satellite status, and improves the ability to perceive the current situation.

[0081] On this basis, potential collision risks are predicted by fusing historical and real-time data based on orbital mechanics equations and gravitational perturbation modeling.

[0082] In step S1, the steps for creating the satellite digital twin are:

[0083] Define track element status: Where r(t) represents the position vector of the satellite at time t, a represents the semi-major axis of the orbit, e represents the orbit eccentricity, and θ(t) represents the true anomaly of the satellite at time t.

[0084] Modeling satellite velocity: Where v(t) represents the velocity of the satellite at time t, μ represents the gravitational parameter of the Earth, r(t) represents the distance from the satellite to the center of the Earth, and a represents the semi-major axis of the orbit;

[0085] The attitude control model is built. The attitude of the satellite is represented by Euler angles, and the attitude change model is represented based on the rigid body dynamics equation. Where I represents the satellite's moment of inertia matrix, ω represents the satellite's angular velocity vector, and T represents the external torque on the satellite;

[0086] Modeling of the near-earth environmental impact: where represents the force caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj represents the projection area of ​​the satellite, which changes dynamically according to the attitude of the satellite n(t), c represents the speed of light, and n(t) represents the normal direction vector of the satellite surface;

[0087] Dynamic modeling of space debris using a distance-based risk model:

[0088] Among them, R collision (t) represents the collision risk at time t, represents the position vector of the satellite at time t, and r debris(t) represents the position vector of the space debris at time t, σ represents the standard deviation of the position uncertainty, which is estimated based on the sensor accuracy and error; the digital twin can track the dynamic state of the satellite in real time and can adjust the orbit and attitude under the influence of external disturbances;

[0089] Specifically, multi-level dynamic dimensionality reduction and compression technology is adopted, and the weight allocation of orbital deviation, attitude adjustment and collision risk dimensions is based on to ensure the prominence of key dynamic dimensions. Dynamic dimensionality reduction based on adaptive weight allocation enables the system to effectively compress high-dimensional data while retaining important information, thereby improving data processing efficiency and reducing information redundancy.

[0090] Step S2, multi-level dimensional compression, extracts important feature data in the space situation based on the digital twin, and dynamically reduces and compresses the multi-dimensional data;

[0091] The twin simulation system extracts features from dynamic data in space situations. Based on historical and real-time data, it identifies and highlights important dynamic change dimensions, including orbital deviation, attitude adjustment, and potential collision risks. It also extracts key information dimensions and uses an adaptive weight allocation algorithm to score the importance of each dimension of data, obtaining multi-dimensional feature data containing dynamic weight allocation results.

[0092] Dynamically select dimensionality reduction algorithm based on multi-dimensional feature data with dynamic weight distribution results;

[0093] In step S2, the dynamic data feature extraction method in the space situation is:

[0094] Extract the characteristics of orbital deviation, Δr(t) = r(t)-r ref (t), where Δr(t) represents the deviation of the satellite from the reference orbital position at time t, r(t) represents the real-time position of the satellite at time t, and r ref (t) represents the reference orbital position based on historical data. The orbital deviation amplitude is in, represents the rate of change of the semi-major axis of the orbit, represents the rate of change of orbital eccentricity;

[0095] Feature extraction for posture adjustment, Δω(t)=ω(t)-ω ref (t), where Δω(t) represents the difference between the real-time angular velocity and the reference angular velocity, ω(t) represents the real-time angular velocity of the satellite at time t, and ω ref (t) represents the reference angular velocity based on historical data;

[0096] The Euler angle change of the posture is, Δφ(t)=φ(t)-φref (t),Δθ(t)=θ(t)-θ ref (t),Δψ(t)=ψ(t)-ψ ref (t), where Δφ(t), Δθ(t), and Δψ(t) represent the differences between the three Euler angles of the satellite attitude and the reference Euler angle, φ(t), θ(t), and ψ(t) represent the Euler angles of the satellite attitude at time t, and φ ref (t),θ ref (t),ψ ref (t) represents the reference Euler angle based on historical data;

[0097] In step S2, the method for extracting dynamic data features in the space situation also includes:

[0098] Extract the features of potential collision risk based on the relative distance and relative speed between the satellite and space debris. rel (t)=||r ref (t)-r debris (t)||, where d rel (t) represents the relative distance between the satellite and the space debris at time t, r ref (t) represents the position vector of the satellite at time t, r debris (t) represents the position vector of the space debris at time t, and the relative velocity is v rel (t)=||v sat (t)-v debris (t)||, where v rel (t) represents the relative speed between the satellite and the space debris, v sat (t) represents the velocity vector of the satellite at time t, v debris (t) represents the velocity vector of the space debris at time t;

[0099] Combined relative distance d rel (t) and relative velocity v rel (t), calculate the probability of collision risk, R collision (t)= Among them, R collision (t) represents the collision risk at time t. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, σ v Represents the velocity attenuation coefficient, calculate R collision (t) Dynamically assess the collision risk between satellites and space debris;

[0100] In step S2, the method for extracting dynamic data features in the space situation also includes:

[0101] Use the adaptive weight assignment algorithm to score the importance of each dimension data, and define the importance weight of the dimension as w i (t), the importance of each dimension at time t is Among them, w i (t) represents the weight of the i-th dimension at time t, Δf i (t) represents the change in the i-th dimension, including orbital deviation, attitude adjustment or collision risk change, and N represents the total number of dynamic dimensions;

[0102] Weight w i (t) Ensure that important dynamically changing dimensions in multidimensional data are highlighted;

[0103] According to the dimension weight w i (t), dynamically adjusted nonlinear dimensionality reduction techniques tSNE and UMAP are selected;

[0104] The cost function is used in the dimensionality reduction process to retain the important information dimensions:

[0105] in, represents the loss function in the dimensionality reduction process, represents the eigenvector of the i-th dimension in high-dimensional space, Represents the feature vector of the i-th dimension in the low-dimensional space, and the loss function is based on the weight w i (t) Dynamic adjustment to effectively extract dynamic features in the space situation and highlight important dimensions (orbit, attitude, collision risk);

[0106] Specifically, by modeling external disturbances to the satellite's orbit and attitude, accurate prediction and dynamic adjustment of disturbances can be achieved, ensuring the accuracy of attitude changes under external disturbance conditions. By dynamically fusing the current situation with future prediction results, rapid response to environmental changes can be achieved, thereby improving the accuracy of space object status tracking and adjustment.

[0107] Step S3, dynamic evolution tracking, dynamic evolution prediction based on the reduced-dimensional data, and integrating the results into visual display;

[0108] Based on the compressed data of step S2, dynamic prediction of future situation is performed, the predicted data is integrated with the current situation information, the dynamic changes of space objects are tracked, and the prediction results of future trajectory and situation are output to provide dynamic evolution data support for the visualization interface, which not only displays the current situation, but also performs dynamic prediction based on twin simulation, effectively tracking the orbit and attitude changes of objects;

[0109] Specifically, the collision risk between space debris and satellites is dynamically calculated by combining relative position and speed. Based on accurate prediction of relative speed and distance changes and control of adaptive attenuation coefficient, the risk assessment model is adjusted in real time and future risks are dynamically updated.

[0110] The dynamic prediction method of future situation based on compressed data in step S3 is:

[0111] Using the compressed data in step S2, combined with the orbit, attitude and potential collision risk key feature dimensions, a dynamic prediction of the future space situation is made, the steps including orbit prediction, attitude prediction and collision risk prediction;

[0112] Orbital dynamics prediction: Where r(t+Δt) represents the position of the satellite at the future time t+Δt, r(t) represents the position of the satellite at the current time t, v(t) represents the velocity of the satellite at the current time t, a(t) represents the acceleration of the satellite at the time t, Δt represents the prediction time step, and the orbital position is predicted using the current velocity v(t) and acceleration a(t), where the acceleration a(t) is modeled by gravity and external disturbances, and the disturbance term is dynamically adjusted based on historical data and simulation models;

[0113] The acceleration of the gravitational disturbance is expressed as Among them, a gravity (t) represents the gravitational acceleration vector, μ represents the earth's gravitational constant, r(t) represents the satellite's current position vector, and the acceleration is calculated by the pull of the earth's gravity on the satellite, and changes dynamically with the position r(t);

[0114] Use rigid body dynamics equations and historical posture data to perform posture dynamics prediction:

[0115] θ(t+Δt)=θ(t)+ω(t)Δt, where θ(t+Δt) represents the attitude angle of the satellite at the future time t+Δt, θ(t) represents the attitude angle of the satellite at the current time t, ω(t) represents the angular velocity vector of the satellite at the current time t, and Δt represents the prediction time step;

[0116] The method of dynamically predicting the future situation based on the compressed data in step S3 also includes:

[0117] To model the attitude disturbance, Among them, T srp (t) represents the torque caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj (t) represents the projected area of ​​the satellite, which depends on the attitude angle θ(t), n(t) represents the normal direction vector of the satellite surface, which depends on the current attitude, c represents the speed of light, and the angular acceleration α(t) caused by the torque is: α(t) = I-1 T srp (t), where α(t) represents the attitude angular acceleration, I represents the satellite’s moment of inertia matrix, and T srp (t) represents the torque caused by the solar radiation pressure, which is numerically integrated and the angular velocity ω(t) is updated using the angular acceleration α(t), thereby updating the satellite's attitude angle θ(t);

[0118] Dynamically predict potential collision risks, based on the relative position and relative speed of the satellite and space debris, predict future collision risks, relative distance d rel (t) and relative speed v rel The evolution of (t) is predicted to be: d rel (t+Δt)=d rel (t)+v rel (t)Δt, where d rel (t+Δt) represents the relative distance between the satellite and the debris at the future time t+Δt, d rel (t) represents the relative distance at the current time t, v rel (t) represents the relative speed at the current time t, and Δt represents the prediction time step;

[0119] Relative speed v at future time rel (t+Δt) is adjusted by the external disturbance prediction model, v rel (t+Δt)=v rel (t)+a rel (t)Δt, where v rel (t+Δt) represents the relative speed at the future time t+Δt, a rel represents relative acceleration;

[0120] Based on the dynamic changes in relative distance and speed, the future collision risk is calculated as follows:

[0121] Among them, R collision (t+Δt) represents the collision risk at the future time t+Δt. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, which controls the weight of relative distance on collision risk, σ v Represents the speed attenuation coefficient, which controls the weight of relative speed on collision risk.

[0122] Integrate dynamic prediction with current situation,

[0123] S future (t+Δt)=λS current (t)+(1-λ)+S predicted (t+Δt), where S future(t+Δt) represents the future situation after fusion, combining current and predicted data, S current (t) represents the current situation data (orbit, attitude, collision risk), S predicted (t+Δt) represents the predicted situation data at the future time, and λ represents the smoothing coefficient, which controls the fusion ratio between the current situation and the future prediction; the current situation is combined with the prediction results to ensure the accuracy of the prediction and the consistency of real-time updates;

[0124] Specifically, through the dimensionality reduction algorithm and dynamic evolution tracking technology in step S2, the multidimensional data is compressed and visualized, focusing on displaying key situation information.

[0125] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A space situation visualization system based on mixed dimensions, characterized by: include, The data acquisition module is responsible for acquiring real-time space situation data and inputting the data into the twin simulation system. The space situation data includes the satellite's orbit information, speed, attitude, near-Earth environment, and dynamic situation information of space debris; The twin simulation system generates a corresponding digital twin based on the real-time data of each satellite, and the digital twin is synchronized with the physical satellite status in real time; Dimensional compression module, which uses the twin simulation system to dynamically extract and compress multi-dimensional data in space situation; Important feature identification module, used to highlight the relevant dimensions of important feature parameters; Dynamic dimensionality reduction module, which dynamically adjusts the dimensionality reduction algorithm based on the prediction results of twin simulation; In the dynamic dimension reduction module, dynamic data feature extraction in the space situation is performed. The extraction method includes: using an adaptive weight allocation algorithm to score the importance of each dimension data, and defining the importance weight of the dimension as w i (t), the importance of each dimension at time t is Among them, w i (t) represents the weight of the i-th dimension at time t, Δf i (t) represents the change in the i-th dimension, including orbital deviation, attitude adjustment or collision risk change, and N represents the total number of dynamic dimensions; According to the dimension weight w i (t), dynamically adjusted nonlinear dimensionality reduction techniques tSNE and UMAP are selected; The cost function is used in the dimensionality reduction process to retain the important information dimensions: in, represents the loss function in the dimensionality reduction process, represents the eigenvector of the i-th dimension in high-dimensional space, Represents the feature vector of the i-th dimension in the low-dimensional space, and the loss function is based on the weight w i (t) Dynamic adjustment; Dynamic evolution tracking and simulation prediction module, which is based on the twin simulation system, performs real-time prediction, tracks the future dynamic changes of the trajectory and attitude of space objects, updates the object status in real time, and integrates the change information into the visualization interface; Visual interface for visual display of multi-dimensional data.

2. A method based on a mixed-dimensional space situation visualization system, characterized in that: Based on the space situation visualization system based on mixed dimensions as described in claim 1, The following steps are included: Step S1, space situation data collection, obtain real-time space situation data from sensors, input the space situation data into the twin simulation system, create a digital twin of each satellite, and synchronize the digital twin with the physical satellite status in real time Step S2, multi-level dimensional compression, extracts important feature data in the space situation based on the digital twin, and dynamically reduces and compresses the multi-dimensional data; Step S3, dynamic evolution tracking, performs dynamic evolution prediction based on the dimensionality reduced data, and integrates the results into visual display.

3. The method based on a mixed-dimensional space situation visualization system according to claim 2, characterized in that: In step S1, the steps for creating the satellite digital twin are: Define track element status: Where r(t) represents the position vector of the satellite at time t, a represents the semi-major axis of the orbit, e represents the orbit eccentricity, and θ(t) represents the true anomaly of the satellite at time t. Modeling satellite velocity: Where v(t) represents the velocity of the satellite at time t, μ represents the gravitational parameter of the Earth, r(t) represents the distance from the satellite to the center of the Earth, and a represents the semi-major axis of the orbit; The attitude control model is built. The attitude of the satellite is represented by Euler angles, and the attitude change model is represented based on the rigid body dynamics equation. Where I represents the satellite's moment of inertia matrix, ω represents the satellite's angular velocity vector, and T represents the external torque on the satellite; Modeling of the near-earth environmental impact: where represents the force caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj represents the projection area of ​​the satellite, which changes dynamically according to the attitude of the satellite n(t), c represents the speed of light, and n(t) represents the normal direction vector of the satellite surface; Dynamic modeling of space debris using a distance-based risk model: Among them, R collision (t) represents the collision risk at time t, represents the position vector of the satellite at time t, and r debris (t) represents the position vector of the space debris at time t, and σ represents the standard deviation of the position uncertainty, which is estimated based on the sensor accuracy and error.

4. The method based on a mixed-dimensional space situation visualization system according to claim 3, characterized in that: In step S2, the twin simulation system extracts features from the dynamic data in the space situation, identifies and highlights important dynamic change dimensions based on historical and real-time data, including orbital deviation, attitude adjustment, and potential collision risks, extracts key information dimensions, and uses an adaptive weight allocation algorithm to score the importance of each dimension of data to obtain multi-dimensional feature data containing dynamic weight allocation results; Dynamically select dimensionality reduction algorithm based on multi-dimensional feature data with dynamic weight assignment results.

5. The method based on a mixed-dimensional space situation visualization system according to claim 4, characterized in that: In step S2, the dynamic data feature extraction method in the space situation is: Extract the characteristics of orbital deviation, Δr(t) = r(t)-r ref (t), where Δr(t) represents the deviation of the satellite from the reference orbital position at time t, r(t) represents the real-time position of the satellite at time t, and r ref (t) represents the reference orbital position based on historical data. The orbital deviation amplitude is in, represents the rate of change of the semi-major axis of the orbit, represents the rate of change of orbital eccentricity; Feature extraction for posture adjustment, Δω(t)=ω(t)-ω ref (t), where Δω(t) represents the difference between the real-time angular velocity and the reference angular velocity, ω(t) represents the real-time angular velocity of the satellite at time t, and ω ref (t) represents the reference angular velocity based on historical data; The Euler angle change of the posture is, Δφ(t)=φ(t)-φ ref (t),Δθ(t)=θ(t)-θ ref (t),Δψ(t)=ψ(t)-ψ ref (t), where Δφ(t), Δθ(t), and Δψ(t) represent the differences between the three Euler angles of the satellite attitude and the reference Euler angle, φ(t), θ(t), and ψ(t) represent the Euler angles of the satellite attitude at time t, and φ ref (t),θ ref (t),ψ ref (t) represents the reference Euler angle based on historical data.

6. The method based on a mixed-dimensional space situation visualization system according to claim 5, characterized in that: In step S2, the method for extracting dynamic data features in the space situation also includes: Extract the features of potential collision risk based on the relative distance and relative speed between the satellite and space debris. rel (t)=||r ref (t)-r debris (t)||, where d rel (t) represents the relative distance between the satellite and the space debris at time t, r ref (t) represents the position vector of the satellite at time t, r debris (t) represents the position vector of the space debris at time t, and the relative velocity is v rel (t)=||v sat (t)-v debris (t)||, where v rel (t) represents the relative speed between the satellite and the space debris, v sat (t) represents the velocity vector of the satellite at time t, v debris (t) represents the velocity vector of the space debris at time t; Combined relative distance d rel (t) and relative speed v rel (t), calculate the probability of collision risk, Among them, R collision (t) represents the collision risk at time t. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, σ V Represents the velocity attenuation coefficient, calculate R collision (t), dynamically assess the collision risk between satellites and space debris.

7. The method based on a mixed-dimensional space situation visualization system according to claim 6, characterized in that: In step S3, a dynamic prediction of the future situation is performed based on the compressed data of step S2, the predicted data is integrated with the current situation information, the dynamic changes of the space object are tracked, and the prediction results of the future trajectory and situation are output.

8. The method based on a mixed-dimensional space situation visualization system according to claim 7, characterized in that: The dynamic prediction method of future situation based on compressed data in step S3 is: Using the compressed data in step S2, combined with the orbit, attitude and potential collision risk key feature dimensions, a dynamic prediction of the future space situation is made, the steps including orbit prediction, attitude prediction and collision risk prediction; Orbital dynamics prediction: Where r(t+Δt) represents the position of the satellite at the future time t+Δt, r(t) represents the position of the satellite at the current time t, v(t) represents the velocity of the satellite at the current time t, a(t) represents the acceleration of the satellite at the time t, Δt represents the prediction time step, and the orbital position is predicted using the current velocity v(t) and acceleration a(t), where the acceleration a(t) is modeled by gravity and external disturbances, and the disturbance term is dynamically adjusted based on historical data and simulation models; The acceleration of the gravitational disturbance is expressed as Among them, a gravity (t) represents the gravitational acceleration vector, μ represents the earth's gravitational constant, and r(t) represents the satellite's current position vector; Use rigid body dynamics equations and historical posture data to perform posture dynamics prediction: θ(t+Δt)=θ(t)+ω(t)Δt, where θ(t+Δt) represents the attitude angle of the satellite at the future time t+Δt, θ(t) represents the attitude angle of the satellite at the current time t, ω(t) represents the angular velocity vector of the satellite at the current time t, and Δt represents the prediction time step.

9. The method based on a mixed-dimensional space situation visualization system according to claim 8, characterized in that: The method of dynamically predicting the future situation based on the compressed data in step S3 also includes: To model the attitude disturbance, Among them, T srp (t) represents the torque caused by solar radiation pressure, P srp represents the solar radiation pressure constant, A proj (t) represents the projected area of ​​the satellite, which depends on the attitude angle θ(t), n(t) represents the normal direction vector of the satellite surface, which depends on the current attitude, c represents the speed of light, and the angular acceleration α(t) caused by the torque is: α(t) = I -1 T srp( t), where α(t) represents the attitude angular acceleration, I represents the satellite's moment of inertia matrix, and T srp (t) represents the torque caused by the solar radiation pressure, which is numerically integrated and the angular velocity ω(t) is updated using the angular acceleration α(t); Dynamically predict potential collision risks, based on the relative position and relative speed of the satellite and space debris, predict future collision risks, relative distance d rel (t) and relative speed v rel The evolution of (t) is predicted to be: d rel (t+Δt)=d rel (t)+v rel (t)Δt, where d rel (t+Δt) represents the relative distance between the satellite and the debris at the future time t+Δt, d rel (t) represents the relative distance at the current time t, v rel (t) represents the relative speed at the current time t, and Δt represents the prediction time step; Relative speed v at future time rel (t+Δt) is adjusted by the external disturbance prediction model, v rel (t+Δt)=v rel (t)+a rel (t)Δt, where v rel (t+Δt) represents the relative speed at the future time t+Δt, a rel represents relative acceleration; Based on the dynamic changes in relative distance and speed, the future collision risk is calculated as follows: Among them, R collision (t+Δt) represents the collision risk at the future time t+Δt. The larger the value, the higher the collision risk. d represents the distance attenuation coefficient, which controls the weight of relative distance on collision risk, σ v represents the speed attenuation coefficient, which controls the weight of relative speed on collision risk. Integrate dynamic prediction with current situation, S future (t+Δt)=λS current (t)+(1-λ)+S predicted (t+Δt), where S future (t+Δt) represents the future situation after fusion, combining current and predicted data, S current (t) represents the situation data at the current moment, S predicted (t+Δt) represents the predicted situation data at the future time, and λ represents the smoothing coefficient, which controls the fusion ratio between the current situation and the future prediction.

Images