Space-based observation configuration optimization method and system based on GDOP optimization

Abstract

The invention discloses a space-based observation configuration optimization method and system based on GDOP optimization. The method comprises the steps that an observation satellite obtains observation information of a target; screening a satellite set capable of participating in observation; enumerating all main and auxiliary satellite combinations meeting task requirements; introducing reinforcement learning, and screening out candidate combinations with high potential; calculating the GDOP value of each candidate combination in parallel in real time; judging and switching the optimal combination; and recording all decision data, efficiency indexes and energy consumption states of the current combination as empirical data. The system comprises a target sensing module, a satellite visibility judgment module, a dynamic resource allocation evaluation module, a main and auxiliary satellite combination enumeration module, a reinforcement learning screening module, a GDOP real-time evaluation module, a configuration switching decision module and a data storage module. According to the invention, accurate real-time positioning can be realized under the condition of dynamic change of the target. The method can be widely applied to the fields of satellite navigation and target observation.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation and target observation, and in particular to a method and system for optimizing space-based observation configurations based on GDOP (Geometry, Difference, and Opposition). Background Technology

[0002] In continuous satellite observation missions of dynamic ground or air targets, positioning accuracy relies heavily on the real-time geometric configuration between the satellite and the target. Its impact is quantified and assessed using the Geometric Dilution of Precision (GDOP). The GDOP value is strictly positively correlated with positioning error: a smaller GDOP value results in higher positioning accuracy and stronger reliability of the measurement results; conversely, an increase in the GDOP value directly leads to a significant amplification of positioning error.

[0003] Traditional space-based observation methods typically employ static mission planning strategies, which fix the primary and secondary satellite configuration before the mission and maintain this configuration throughout the mission. Such methods fail to respond to changes in relative geometry caused by target maneuvers and space environment disturbances, exhibiting significant limitations, including: the GDOP value of the original satellite configuration deteriorates rapidly when the target maneuvers quickly or is affected by orbital perturbations; maintaining basic observation capabilities requires a large number of redundant satellite backups, resulting in low resource utilization (typically less than 40%); and once the configuration fails, the lack of real-time autonomous reconfiguration capabilities easily leads to target loss and observation interruption. Summary of the Invention

[0004] In view of this, in order to solve the technical problem of low information resource utilization caused by fixed configuration in existing satellite observation methods, firstly, this invention proposes a space-based observation configuration optimization method based on GDOP optimization, which includes: First, the observation satellite acquires and processes the target's observation information (such as azimuth and elevation angles) in real time through multi-source data fusion (such as information from multiple sensors including optics, infrared, and radar), thereby comprehensively perceiving the target's status. All raw data is processed and preprocessed locally by each satellite node under a distributed edge computing architecture, significantly reducing the load on the central node and improving response speed.

[0005] Then, based on the current visibility of satellites and targets, and combined with dynamic resource allocation and energy consumption optimization strategies, a set of satellites suitable for observation is selected. This strategy comprehensively evaluates the remaining energy, payload status, and link quality of satellites, prioritizing satellites with high health and balanced energy consumption to form the candidate set.

[0006] Next, all primary and secondary star combinations that meet the task requirements are enumerated (e.g., one primary and three secondary, two primary and multiple secondary, etc.). During this process, reinforcement learning is introduced to optimize GDOP prediction and switching decisions. Instead of blindly enumerating all combinations, the system uses a prediction model trained with historical performance data to intelligently pre-screen high-potential candidate combinations, significantly reducing the amount of computation.

[0007] Subsequently, the GDOP value of each candidate combination is calculated in parallel and in real time based on distributed computing nodes, and the results are sorted.

[0008] Next, the optimal combination is determined and switched. The decision engine comprehensively considers the GDOP value, the energy cost of switching, and the collaborative state of the distributed nodes. If the prediction model determines that the GDOP value of the current combination is poor, and there is a better combination, and the switching threshold condition is met (such as the performance improvement benefit being higher than the energy cost of communication and attitude adjustment), then the primary and secondary satellite roles are switched; if there is no better combination or the switching cost is too high, the current configuration is maintained.

[0009] Then, using the observation information of the selected primary and secondary satellite combination, and through multi-source data fusion technology (fusing external information such as inter-satellite ranging and AIS signals), high-precision positioning of the observation target is achieved.

[0010] Finally, all decision data, performance indicators, and energy consumption status of the current combination are recorded as empirical data in a distributed database for continuous training and optimization of the reinforcement learning model. Simultaneously, the system sets a minimum hold period and dynamically adjusts this period based on reinforcement learning to avoid frequent switching that could lead to system instability, thus achieving an optimal balance between long-term performance and energy consumption.

[0011] Secondly, in addition to the above-mentioned space-based observation configuration optimization method based on GDOP optimization, a space-based observation configuration optimization system based on GDOP optimization is also proposed, including: a target perception module, a satellite visibility determination module, a dynamic resource allocation evaluation module, a primary and secondary satellite combination enumeration module, a reinforcement learning screening module, a GDOP real-time evaluation module, a configuration switching decision module, and a data storage module.

[0012] Based on the above scheme, this invention provides a method and system for optimizing space-based observation configurations based on optimal GDOP (Government Doppler Optical Target). This overcomes the problems of lack of dynamic adaptability, low utilization of measured resources, and delayed failure response encountered by traditional observation modes that employ static mission planning. When the target changes dynamically, the optimal GDOP is selected in real time using reinforcement learning, and the primary and secondary satellite combinations are switched, thereby achieving accurate real-time positioning of non-cooperative targets and compensating for the shortcomings of traditional systems. Attached Figure Description

[0013] Figure 1This is a flowchart of the steps of the space-based observation configuration optimization method based on GDOP of the present invention; Figure 2 This is a logical schematic diagram of the main-auxiliary star configuration switching in a specific embodiment of the present invention; Figure 3 This is a comparative diagram of the method of the present invention and the traditional static GDOP method; Figure 4 This is a structural block diagram of a space-based observation configuration optimization system based on GDOP optimization according to the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] It should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0016] It should be understood that the terms "system," "apparatus," "unit," and / or "module" used in this application are a method of distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0017] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "a," and / or "the" are not specifically singular and may include the plural. Generally, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements. An element defined by the phrase "comprising an..." does not exclude the presence of other identical elements in the process, method, product, or apparatus that includes the element.

[0018] In the description of the embodiments of this application, "a plurality of" refers to two or more. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0019] Furthermore, flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, the steps can be processed in reverse order or simultaneously. Additionally, other operations can be added to these processes, or one or more steps can be removed from them.

[0020] Reference Figure 1 This is a flowchart illustrating an optional example of the space-based observation configuration optimization method based on GDOP proposed in this invention. This method can be applied to computer equipment. The space-based observation configuration optimization method proposed in this embodiment may include, but is not limited to, the following steps: Step S1: Perceive the target and acquire target observation information; Step S2: Based on the satellite orbits and the target observation information, construct a list of visible satellites; Step S3: Based on the preset quantity requirements, construct a set of primary and secondary satellite combination schemes according to the visible satellite list; Step S4: Based on the pre-trained reinforcement learning model, the performance of the primary and secondary satellite combination scheme set is predicted and filtered to obtain the filtered satellite combination list; Step S5: Perform real-time GDOP evaluation on the filtered satellite combination list to obtain the corresponding real-time GDOP value; Step S6: Select the optimal primary and secondary star combination and perform configuration switching based on the preset threshold strategy and real-time GDOP value.

[0021] In some feasible embodiments, step S1 specifically includes: Collect the target's physical state information (relative azimuth and elevation angles) to provide basic data for subsequent observation and configuration decisions.

[0022] The main technical components include sensor fusion, dynamic modeling, and uncertainty handling. Sensor fusion technology refers to the fusion of multiple sensor sources such as ground radar, infrared detection, and space telescopes to improve perception accuracy and robustness. Dynamic modeling technology refers to predicting the next state of the target based on historical trajectories and kinematic models (such as two-body motion models). Uncertainty handling refers to processing the uncertainty of the target's state through Bayesian estimation or Kalman filtering to improve system robustness. The output includes the target's real-time position, velocity, acceleration, and other states, as well as state confidence or covariance matrix.

[0023] In some feasible embodiments, step S2 specifically includes: Determining which satellites can observe the target at the current or predicted time, based on satellite orbits and target status, is a preliminary screening step for subsequent combination enumeration.

[0024] The main technical components include: orbit prediction, visibility geometry assessment, and dynamic temporal evaluation. Orbit prediction refers to satellite orbit propagation using SGP4 or high-precision numerical integration; visibility geometry assessment determines whether current conditions meet visibility requirements (e.g., elevation angle > 10°, no obstructions, and an observation window exists); dynamic temporal evaluation supports prediction of visibility changes over the next few minutes or hours for advance planning. The output includes a list of visible satellites at the current time and within the predicted time period, as well as an observation condition score or visibility score for each satellite.

[0025] In some feasible embodiments, after step S2, the method further includes: Based on satellite platform status, energy constraints, and link conditions, the optimal subset of missions to be executed is selected from available satellites, providing high-quality input for subsequent combination and generation.

[0026] The main technical components include: multi-constraint optimization modeling, resource efficiency assessment, and dynamic weight allocation. Multi-constraint optimization modeling refers to establishing a function aimed at maximizing mission lifetime and system robustness, with constraints including satellite remaining energy, payload health status, and inter-satellite link quality. Resource efficiency assessment involves calculating the execution efficiency score for each satellite using a comprehensive scoring algorithm (such as a weighted scoring model or heuristic algorithm). Dynamic weight allocation refers to adjusting the weight coefficients in the assessment criteria (such as dynamic adjustment of energy consumption weight and accuracy weight) in real time according to mission priority. The output is a list of optimized and screened candidate satellites and their comprehensive efficiency scores.

[0027] The formula for calculating the performance score is as follows: in, Indicates the first The performance score of each satellite. Indicates the first Remaining energy score of each satellite Indicates the first Payload health status score for each satellite Indicates the first Inter-satellite link quality score for each satellite express, , , and This represents the corresponding weighting coefficient.

[0028] Adaptive weight calculation based on task priority: in, Indicates time Energy weighting coefficient, Indicates the base weight value. This indicates the priority factor for task energy requirements. This represents the system's energy margin assessment factor. , This represents the corresponding adjustment coefficient.

[0029] Configuration robustness index calculation: in, This represents the system's robustness score. The coefficient of variation for performance rating. For the worst link quality, Number of available backup satellites , and This represents the robustness weighting coefficient for the beverage.

[0030] Maximize task lifespan calculation: in, Indicates the expected mission lifespan. Indicates the first The mission weight of each satellite Indicates the first Energy consumption rate of each satellite This represents the robustness impact coefficient.

[0031] Optimize the calculation of the filtering threshold: in, Indicates the performance rating threshold. This represents the average performance score. Indicates the standard deviation of performance ratings. This represents the adjustment factor, which is usually between 1 and 2.

[0032] In some feasible embodiments, step S3 specifically includes: Based on the given satellite quantity requirement, all valid primary and secondary satellite combinations are systematically constructed from the candidate satellite list, providing a basic set of schemes for accuracy assessment. The main technical components include: combinatorial mathematical modeling, spatial geometry optimization, and heuristic search strategies. Combinatorial mathematical modeling refers to applying the combinatorial principle to calculate all possible combinations of selecting k satellites from n satellites; spatial geometry optimization prioritizes satellite spatial distribution characteristics (such as baseline length and geometric configuration diversity) during enumeration; and heuristic search strategies employ heuristic rules (such as genetic algorithms and greedy algorithms) to reduce computational complexity when the combination scale is too large. The output is a complete set of schemes containing all possible primary and secondary satellite combinations.

[0033] In some feasible embodiments, step S4 specifically includes: By using machine learning to intelligently select high-potential satellite combinations and predicting the expected performance of different combinations through pre-trained reinforcement learning models, the computational scale can be significantly reduced and the system efficiency improved.

[0034] The main technical components include: state feature engineering, reward function design, and decision strategy optimization. State feature engineering refers to abstracting the characteristics of satellite combinations (such as satellite geometric distribution and historical performance data) into state vectors that can be processed by machine learning; reward function design refers to constructing a reward mechanism based on indicators such as GDOP value and energy consumption cost to guide the agent to learn the optimal selection strategy; decision strategy optimization refers to continuously optimizing the selection strategy through deep Q-networks (DQN) or policy gradient methods. The output is a short list of pre-selected high-potential satellite combinations.

[0035] In some feasible embodiments, step S5 specifically includes: The precise calculation of the geometric accuracy factor of the satellite combination is achieved through a parallel computing architecture, which quickly evaluates the theoretical positioning accuracy of each combination and provides a quantitative basis for decision-making.

[0036] The main technical aspects include: geometric matrix construction, matrix operation optimization, and parallel computing acceleration. Geometric matrix construction refers to constructing the observation geometric matrix H based on the relative positions of the satellite and the target; matrix operation optimization refers to using numerically stable algorithms to solve matrix inverses and trace operations to ensure computational accuracy; parallel computing acceleration refers to utilizing hardware such as GPUs / FPGAs to perform parallel processing of GDOP calculations for multiple combinations. A sorted list of GDOP values ​​for all candidate combinations is output.

[0037] GDOP is calculated based on the Jacobian matrix and covariance of the measured angles between the observed satellite and the target.

[0038] In some feasible embodiments, step S6 specifically includes: Analyze whether the current portfolio needs to be changed (e.g., if the current GDOP is rising too quickly), and make switching decisions based on the recommended new portfolio to achieve real-time adaptability.

[0039] The main technical aspects include: switching condition judgment, balancing switching costs, and state preservation strategies. Switching condition judgment refers to setting switching trigger thresholds, such as GDOP thresholds, observation failures, and target mutations; balancing switching costs involves considering observation interruptions, controlling resource consumption, and avoiding frequent switching; the state preservation strategy means maintaining the status quo if the current configuration still meets the requirements, thus reducing resource overhead. The output includes the switching decision, the new master-slave combination configuration, and the switching command, which is then sent to the observation task system.

[0040] in, Figure 2This is the control logic diagram for master-slave satellite configuration switching. This logic achieves intelligent switching decisions through dual condition constraints: first, it checks whether the change in the current configuration's GDOP value exceeds a set threshold; if the threshold condition is met, it further determines whether the current configuration's duration has reached the minimum hold time T_min. Configuration switching is only performed when both the GDOP change condition and the time stability condition are met simultaneously; otherwise, it enters a hold or delayed waiting state. This design effectively suppresses system oscillations caused by frequent switching, achieving a balance between positioning accuracy optimization and system stability.

[0041] In some feasible embodiments, it also includes: Step S7, data storage.

[0042] Responsible for the persistent storage and management of data throughout the system, recording system operating status, decision data, and performance indicators to provide a data foundation for system optimization.

[0043] The main technical components include: time-series data management, distributed storage architecture, and data service interfaces. Time-series data management refers to the efficient storage and retrieval of timestamped process data using a time-series database; the distributed storage architecture refers to the reliable storage of massive amounts of data using a distributed file system (such as HDFS) or database (such as Cassandra); and the data service interface refers to providing standard APIs for other modules to query historical data and support machine learning training. The output consists of a structured, full-process data package and a model training dataset.

[0044] This method offers significant advantages in task execution, including high accuracy, timely response, good stability, strong security, and high flexibility, while greatly reducing maintenance and upgrade difficulties. This novel technology not only solves the problems of traditional systems but also brings significant improvements in task execution, making it more reliable in dynamic and complex environments.

[0045] Based on the above method and process, Figure 3 The diagram illustrates a comparison between the traditional static GDOP method and the dynamic GDOP reselection method of this invention. This comparison clearly reveals the technical differences between the traditional static GDOP method and the dynamic GDOP reselection method of this invention from five core dimensions: configuration stability, primary and secondary satellite roles, long-term GDOP performance, satellite resource utilization, and configuration failure response. Traditional methods, relying on initial design, suffer from poor configuration stability, fixed primary and secondary satellite roles, easy GDOP degradation, redundant satellite resource deployment, and a lack of adaptive capabilities. In contrast, this invention achieves dynamic configuration optimization through real-time evaluation and adjustment, supports dynamic switching of primary and secondary satellite roles to maintain long-term near-optimal GDOP performance, improves satellite resource utilization through combinatorial optimization scheduling, and possesses the adaptive response capability for real-time reconstruction of observation configurations, significantly improving the technical shortcomings of traditional methods.

[0046] like Figure 4 As shown, a space-based observation configuration optimization system based on GDOP optimality includes: The target perception module is used to execute step S1; The satellite visibility determination module is used to execute step S2; The dynamic resource allocation and evaluation module selects the optimal subset of missions from available satellites based on satellite platform status, energy constraints, and link conditions. The primary and secondary star combination enumeration module is used to execute step S3; The reinforcement learning selection module is used to perform step S4; The GDOP real-time evaluation module is used to perform step S5; Configuration switching decision module, used to execute step S6; The data storage module is used to execute step S7.

[0047] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0048] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A GDOP-optimization-based space-based observation configuration optimization method, characterized in that, The method comprises the following steps: perceiving a target and obtaining target observation information; constructing a visible satellite list based on satellite orbits and the target observation information; constructing a master-satellite combination scheme set based on the visible satellite list and a preset quantity requirement; performing performance prediction and screening on the master-satellite combination scheme set based on a pre-trained reinforcement learning model to obtain a screened satellite combination list; performing real-time GDOP evaluation on the screened satellite combination list to obtain corresponding real-time GDOP values; selecting an optimal master-satellite combination and performing configuration switching according to a preset threshold strategy and the real-time GDOP values.

2. The GDOP-optimal space-based observation configuration optimization method according to claim 1, further comprising: optimizing the visible satellite list based on satellite platform states, energy constraints and link conditions. 3.The method of claim 1, wherein, Further comprising: saving records of configuration switching as experience data; adjusting a shortest maintenance period of the configuration based on the experience data and the reinforcement learning model.

4. The method of claim 1, wherein the GDOP-optimized space-based observation configuration is optimized based on a GDOP value. The preset threshold strategy comprises: triggering a configuration switching instruction when a real-time GDOP value drop ratio exceeds a preset threshold and a configuration maintenance time of the satellite combination reaches a minimum maintenance period.

5. The method of claim 1, wherein the GDOP-optimized space-based observation configuration is optimized based on a GDOP value. The step of constructing a master-satellite combination scheme set based on the visible satellite list and a preset quantity requirement specifically comprises: selecting all possible combination modes of a preset number of satellites from the visible satellite list by considering satellite spatial distribution characteristics based on combination number principles; when a size of the all possible combination modes is greater than a preset value, reducing calculation complexity by using heuristic rules to obtain the master-satellite combination scheme set.

6. The method of claim 1, wherein the GDOP-optimized space-based observation configuration is optimized based on a GDOP value. The step of performing real-time GDOP evaluation on the screened satellite combination list to obtain corresponding real-time GDOP values specifically comprises: constructing an observation geometry matrix based on the screened satellite combination list according to relative positions of satellites and targets; solving matrix inverse and trace operations by using a numerical stability algorithm, and performing parallel processing on GDOP calculations of multiple combinations to obtain real-time GDOP values of corresponding satellite combinations.

7. The method of claim 1, wherein the GDOP-optimized space-based observation configuration is determined based on a GDOP value of each of the plurality of space-based observation configurations. The step of optimizing the visible satellite list based on satellite platform states, energy constraints and link conditions specifically comprises: calculating an execution performance score of each satellite by considering residual energy of the satellite, a load health state and inter-satellite link quality through a comprehensive scoring algorithm, with maximization of task life and system robustness as objective functions; adjusting weight coefficients in the comprehensive scoring algorithm in real time according to task priorities; performing optimization screening based on the execution performance score and a screening threshold to obtain a candidate satellite list.

8. The method of claim 7, wherein the GDOP-optimal space-based observation configuration is determined by solving the following optimization problem: min x∈X GDOP(x) subject to x∈X, where X is a set of feasible configurations. A calculation formula of the execution performance score of the satellite is as follows: wherein, represents the performance score of the i-th satellite, represents the residual energy score of the i-th satellite, represents the payload health score of the i-th satellite, represents the inter-satellite link quality score of the i-th satellite, represents the system redundancy score of the i-th satellite, and represents the corresponding weight coefficient.​​​​​​​ 9.A GDOP-optimal space-based observation configuration optimization system, characterized in that, comprises: a target perception module; a target perception module; performing real-time GDOP evaluation on the screened satellite combination list to obtain corresponding real-time GDOP values; selecting an optimal master-satellite combination and performing configuration switching according to a preset threshold strategy and the real-time GDOP values; a satellite visibility determination module, which constructs a visible satellite list based on satellite orbits and the target observation information; a dynamic resource allocation evaluation module, which optimizes the visible satellite list based on satellite platform state, energy constraints and link conditions; a primary and secondary satellite combination enumeration module, which constructs a primary and secondary satellite combination scheme set according to the visible satellite list based on a preset quantity requirement; a reinforcement learning screening module, which performs performance prediction and screening on the primary and secondary satellite combination scheme set based on a pre-trained reinforcement learning model to obtain a screened satellite combination list; a GDOP real-time evaluation module, which performs GDOP real-time evaluation on the screened satellite combination list to obtain corresponding real-time GDOP values; a configuration switching decision module, which selects an optimal primary and secondary satellite combination and performs configuration switching according to a preset threshold strategy and the real-time GDOP values; a data storage module, which saves records of configuration switching as experience data.

Images