A pipeline control method, device, medium and equipment for a satellite retrograde orbit

Abstract

The application discloses a kind of pipeline control methods, devices, media and equipment of satellite regression orbit, belong to satellite control field.The actual satellite and reference orbit data are obtained first in the application, and the spatial error vector is calculated in combination with the time alignment strategy along the track;When the orbit needs to be maintained, the theoretical tangential impulse and the best phase are obtained based on the actual elements using the deterministic reference model;Then, using the impulse as a reference, introduce uncertainty parameters, use Latin hypercube sampling and NSGA-III double-layer algorithm to solve the Pareto optimal solution set;Select the robust optimal impulse from it to generate instructions, control satellite maneuvering, and realize the pipeline control of regression orbit.The application effectively solves the problem that the existing technology cannot accurately and efficiently control the pipeline of satellite regression orbit.

Description

Technical Field

[0001] This invention relates to the field of satellite control, and in particular to a pipeline control method, apparatus, medium and equipment for satellite reentry into orbit. Background Technology

[0002] A strictly recurring orbit for a low-Earth orbit (LEO) satellite refers to a special type of orbit in which the satellite's nadir trajectory repeats strictly within a fixed period, and the spacing between adjacent trajectories remains constant. This type of orbit is widely used in Earth observation missions that demand extremely high stability of the spatiotemporal baseline, such as interferometric synthetic aperture radar (IASAR), surface deformation monitoring, and digital elevation measurements. To ensure high coherence and high-quality imaging in interferometry, the satellite's actual trajectory must always remain within a virtual space conduit of a given radius near the reference orbit. This requires a control strategy with extremely high precision, capable of simultaneously overcoming the semi-major axis decay caused by atmospheric drag and orbital plane drift caused by the Earth's non-spherical gravity.

[0003] Currently, for pipeline maintenance in strictly regressive orbits, engineering primarily employs predictive control strategies based on deterministic models, such as those used by the Japanese ALOS-2 satellite and the German TerraSAR-X satellite. This type of existing technology mainly predicts the future state of the satellite using a high-precision orbital predictor under an ideal dynamic model, and employs a dead-zone threshold strategy. This means that when the predicted deviation is about to exceed a safety threshold set within the pipeline boundary or interior, an orbital maneuver is triggered. Its core is based on deterministic optimization algorithms, which, assuming the actuators can perfectly execute commands, find an optimal velocity increment that maximizes the time interval between the next maneuver or minimizes fuel consumption without violating pipeline constraints.

[0004] However, the aforementioned existing technologies have significant drawbacks in practical engineering applications. First, existing technologies ignore the uncertainties of the actuators, resulting in poor robustness. Existing methods typically assume that the ignition timing, thrust amplitude, and thrust direction of pulse maneuvers can be executed precisely. However, in actual on-orbit operation, due to thruster performance limitations and attitude control accuracy constraints, thrust amplitude deviations, pointing errors, and ignition time delays are inevitable. Existing technologies cannot eliminate the cumulative effects of these random deviations in the control loop beforehand. Second, existing technologies have low duct constraint satisfaction rates and high failure risks. Because the deterministic optimal solution aims to maximize maintenance time, its orbital error peak is usually close to the duct edge, and the system lacks the necessary safety margin. Once a small execution deviation occurs, the actual orbit is very likely to diverge and exceed the duct radius limit. Finally, existing technologies lack a performance-stability trade-off mechanism and cannot find the optimal balance between maintenance benefits and out-of-bounds risks based on error distribution characteristics. This often leads to control strategies that are either too aggressive, resulting in out-of-bounds errors, or too conservative, resulting in frequent maneuvers and wasted fuel. These shortcomings prevent existing technologies from accurately and efficiently controlling the duct of a satellite's return orbit. Summary of the Invention

[0005] This invention provides a method, apparatus, medium, and equipment for controlling the pipeline of a satellite returning to its orbit, in order to solve the problem that the pipeline of a satellite returning to its orbit cannot be accurately and efficiently controlled in the prior art.

[0006] Firstly, this application provides a pipeline control method for satellite reentry orbit, including: Acquire actual satellite orbit measurement data and preset reference orbit data; Based on the actual satellite orbit measurement data and reference orbit data, and combined with the preset time alignment strategy along the track, the satellite orbit spatial error vector is calculated. Based on the satellite orbital spatial error vector, it is determined whether orbit maintenance is required. If orbit maintenance is required, the current orbital elements are determined based on the actual orbital measurement data, and the orbital elements are input into the preset deterministic benchmark maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase. Based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, combined with the preset two-layer solution algorithm based on Latin hypercube sampling and third-generation non-dominated sorting genetic algorithm, the Pareto optimal solution set is calculated. The robust optimal pulse amplitude is selected from the Pareto optimal solution set, and control commands are generated based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, thereby achieving pipeline control of the return orbit.

[0007] This application first acquires actual satellite orbit measurement data and reference orbit data, and then accurately calculates the spatial error vector using a time alignment strategy along the track, eliminating the error assessment bias caused by neglecting phase drift in traditional methods. When it is determined that orbit maintenance is necessary, a deterministic benchmark strategy is used to quickly obtain the theoretical tangential pulse amplitude and optimal maneuver phase based on actual orbit elements, providing a reliable nominal benchmark for subsequent optimization. Then, using this theoretical pulse amplitude as the search center, uncertainty parameters of the actuator are introduced, and a two-layer solution algorithm based on Latin hypercube sampling and a third-generation non-dominated sorting genetic algorithm is employed for uncertainty propagation and multi-objective optimization. This maximizes the expected performance of the effective maintenance orbits while minimizing the performance fluctuation variance. Finally, the robust optimal pulse amplitude is selected from the Pareto optimal solution set to execute the orbit maneuver. This allows the control system to explicitly quantify the impact of thrust amplitude, direction, and time deviations, significantly reducing the risk of exceeding limits due to execution errors. It achieves an optimal trade-off between maintenance benefits and system stability, effectively improving the reliability and robustness of strict return-to-orbit pipeline control for satellites in non-ideal environments. This application effectively solves the problem that existing technologies cannot accurately and efficiently control the pipeline for satellite return-to-orbit.

[0008] Furthermore, the calculation of the satellite orbit spatial error vector based on the actual satellite orbit measurement data and reference orbit data, combined with a preset time alignment strategy along the track, specifically involves: Establish radial, tangential, and normal orbital coordinate systems at the reference orbit sampling points; Based on the phase drift along the track direction caused by atmospheric drag on the actual orbit, within the time window of the reference orbit sampling point, a search algorithm is used to find the alignment moment that minimizes the tangential position deviation between the actual orbit and the reference orbit in the radial, tangential and normal orbit coordinate systems. Based on the alignment time, the position vector of the actual orbit is subtracted from the position vector of the reference orbit, and a coordinate transformation is performed using the rotation matrix from the preset geocentric coordinate system to the radial, tangential, and normal orbit coordinate systems to calculate the satellite orbit spatial error vector after eliminating the time phase difference.

[0009] This application establishes radial, tangential, and normal orbital coordinate systems at the sampling point of the reference orbit. Based on the phase drift along the track direction caused by atmospheric drag on the actual orbit, it searches for the alignment moment with the smallest tangential position deviation within a time window near the sampling point. This eliminates the huge tangential error caused by phase drift when directly comparing positions at the same moment in traditional methods. Then, based on this alignment moment, the position vectors of the actual orbit and the reference orbit are subtracted, and a coordinate transformation is performed using a rotation matrix from the geocentric Earth-fixed coordinate system to the orbital coordinate system to accurately calculate the spatial error vector after eliminating the time phase difference. This process avoids misjudgments caused by phase drift and provides a reliable geometric basis for subsequent pipeline constraint satisfaction judgments, making the assessment of whether the satellite has exceeded its limits more accurate. This effectively reduces the risk of unnecessary maneuvers or exceeding limits caused by assessment deviations, significantly improving the accuracy and reliability of strict regression orbit pipeline control.

[0010] Furthermore, the step of determining the current orbital elements based on the actual orbital measurement data and inputting the orbital elements into a preset deterministic benchmark maintenance strategy model to obtain the theoretical tangential pulse amplitude and optimal maneuver phase specifically involves: Based on the difference between the frozen eccentricity vector in the reference orbit data and the eccentricity vector in the actual orbit measurement data, the latitude argument used to determine the optimal ignition phase is calculated, and the latitude argument is taken as the optimal maneuver phase. With the optimization objective of maintaining the peak value of the normal error within a preset period close to the preset pipeline boundary, an objective function is constructed, and the Newton-Raphson iteration method is used to solve the objective function to obtain the theoretical tangential pulse amplitude.

[0011] This application first calculates the latitude argument used to determine the optimal ignition phase based on the difference between the frozen eccentricity vector of the reference orbit and the eccentricity vector of the actual orbit. This ensures that the tangential pulse can be applied at the most advantageous orbital position, avoiding a decrease in control efficiency or the introduction of additional disturbances due to improper phase. Based on this, an objective function is constructed with the goal of maintaining the peak normal error close to the preset pipe boundary within a preset period. The Newton-Raphson iteration method is then used to solve this function, yielding the theoretical tangential pulse amplitude. This process fully utilizes frozen orbit theory and numerical iterative optimization, enabling the acquisition of a long-period sustained solution that keeps the orbital error close to the pipe boundary under ideal error-free conditions. This provides a high-reference nominal control quantity for subsequent robust optimization, ensuring both control economy and laying a solid foundation for finding the optimal solution that balances performance and stability under uncertain environments.

[0012] Furthermore, based on the theoretical tangential pulse amplitude and preset uncertainty parameters, combined with a preset two-layer solution algorithm based on Latin hypercube sampling and third-generation non-dominated sorting genetic algorithm, the Pareto optimal solution set is calculated, specifically as follows: The theoretical tangential pulse amplitude and the preset uncertainty parameters are input into a two-layer solution algorithm for uncertainty propagation and multi-objective optimization. In the outer loop, the initial population of individuals with each tangential pulse amplitude is generated using the third-generation non-dominated sorting genetic algorithm. Each individual in the initial population is passed to the inner loop. In the inner loop, the uncertainty parameter is stratified and sampled using Latin hypercube sampling to generate a physical error sample set. The physical error sample set is then injected into a preset orbital recursion model to evaluate the statistical characteristics of the number of maintenance cycles of the current individual under error interference. The inner loop passes the mean and variance of the number of maintenance cycles obtained from the evaluation to the outer loop as the optimization target value for the individual. The outer loop performs non-dominated sorting and selection on all individuals according to the target value, generates a new generation of population, and iterates until the preset termination condition is met, outputting the Pareto optimal solution set.

[0013] This application employs a two-layer solution algorithm by inputting theoretical tangential pulse amplitude and preset uncertainty parameters. The outer layer uses a third-generation non-dominated sorting genetic algorithm to generate a population of individuals with tangential pulse amplitude values. The inner layer uses Latin hypercube sampling to perform stratified sampling of the uncertainty parameters and injects them into the orbital recursive model. This efficiently and accurately evaluates the mean and variance of the number of maintenance cycles for each individual under error disturbances. The inner layer returns the mean and variance to the outer layer as optimization objectives. The outer layer drives population evolution through non-dominated sorting and selection, ultimately outputting a Pareto optimal solution set. This process enables the control system to explicitly quantify the impact of thrust amplitude, direction, and time deviations on orbital maintenance performance, minimizing performance fluctuations while maximizing the expected number of maintenance cycles. This achieves an optimal trade-off between maintenance efficiency and robust stability, significantly reducing the risk of out-of-bounds errors caused by execution errors and effectively solving the problems of low constraint satisfaction and high failure risk in pipelines under non-ideal environments using traditional deterministic control strategies.

[0014] Furthermore, the preset uncertainty parameters include thrust amplitude error, thrust direction error, and execution timing error; wherein, the thrust amplitude error is used to simulate the deviation of the actual thrust magnitude from the preset command thrust, the thrust direction error is modeled using a preset conical distribution model, and the execution timing error is used to simulate the ignition position deviation caused by clock drift or command delay.

[0015] This application incorporates thrust amplitude error, thrust direction error, and execution timing error into a pre-defined uncertainty parameter system. Thrust amplitude error simulates the random deviation of the actual thrust from the commanded thrust; thrust direction error is modeled using a conical distribution model to characterize the degree to which the actual thrust vector deviates from the nominal tangential axis in three-dimensional space; and execution timing error simulates ignition position deviation caused by clock drift or command delay. This allows for the first explicit and comprehensive quantification of actuator uncertainties in the orbital control loop. This modeling process enables subsequent two-layer solution algorithms to perform uncertainty propagation and multi-objective optimization based on the real error distribution, rather than relying on an ideal error-free assumption. This allows for the pre-assessment and suppression of the cumulative effects of various deviations during the optimization phase. Consequently, the control system can find the optimal balance between maintaining efficiency and overshoot risk, significantly improving the pipeline constraint satisfaction rate in non-ideal environments and effectively solving the problems of poor robustness and high failure risk caused by neglecting actuator uncertainties in existing technologies.

[0016] Secondly, this application provides a pipeline control device for satellite reentry into orbit. The pipeline control device for satellite reentry into orbit includes: The acquisition module is used to acquire actual satellite orbit measurement data and preset reference orbit data; The first calculation module is used to calculate the satellite orbit spatial error vector based on the actual orbit measurement data and reference orbit data of the satellite, combined with a preset time alignment strategy along the track. The processing module is used to determine whether orbit maintenance is required based on the satellite orbital spatial error vector. If orbit maintenance is required, it determines the current orbital elements based on the actual orbital measurement data and inputs the orbital elements into a preset deterministic reference maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase. The second calculation module is used to calculate the Pareto optimal solution set based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, combined with the preset two-layer solution algorithm based on Latin hypercube sampling and the third-generation non-dominated sorting genetic algorithm. The control module is used to select the robust optimal pulse amplitude from the Pareto optimal solution set, and generate control commands based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, so as to realize pipeline control of the return orbit.

[0017] This application acquires actual satellite orbit measurement data and reference orbit data through an acquisition module. A first calculation module, combined with a time alignment strategy along the track, accurately calculates the spatial error vector, eliminating evaluation bias caused by phase drift. A processing module determines whether orbit maintenance is necessary based on the error vector. When needed, it uses a deterministic baseline maintenance strategy model to quickly obtain the theoretical tangential pulse amplitude and optimal maneuver phase, providing a reliable nominal baseline for optimization. A second calculation module, centered on this theoretical pulse amplitude, introduces uncertainty parameters and employs a two-layer solution algorithm based on Latin hypercube sampling and a third-generation non-dominated sorting genetic algorithm for multi-objective optimization, generating a Pareto optimal solution set that balances desired maintenance performance and robust stability. Finally, a control module selects the robust optimal pulse amplitude from this set to generate commands for orbit maneuvering. The modules work together and the data flow is clear, enabling the control system to explicitly quantify the impact of thrust amplitude, direction and time deviation. While maximizing the expected number of orbits, it minimizes performance fluctuations, significantly reduces the risk of exceeding limits caused by execution errors, achieves the optimal trade-off between maintenance benefits and system stability, and effectively improves the reliability and robustness of strict return-to-orbit pipeline control of satellites in non-ideal environments.

[0018] Thirdly, this application provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform a pipeline control method for satellite reentry into orbit as described above. Its beneficial effects are the same as those of the pipeline control method for satellite reentry into orbit provided in the first aspect of this application.

[0019] Fourthly, this application provides a terminal device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement any of the satellite return-to-orbit pipeline control methods described in the first aspect. Attached Figure Description

[0020] Figure 1 : A schematic flowchart of an embodiment of the pipeline control method for satellite return orbit provided in this application; Figure 2 : A schematic diagram of an embodiment of the orbital spatial error provided in this application; Figure 3 : A schematic diagram of an embodiment of the conical distribution model provided in this application; Figure 4 : A schematic diagram of an embodiment of the algorithm flow provided in this application; Figure 5 : A schematic diagram of an embodiment of the RN plane error distribution comparison provided in this application; Figure 6 : A schematic diagram of an embodiment of the pipeline control device for satellite return to orbit provided in this application. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1 Please refer to Figure 1 In order to solve the problem that existing technologies cannot accurately and efficiently control the pipeline of a satellite returning to its orbit, this invention provides a pipeline control method for a satellite returning to its orbit, including steps S01-S05.

[0023] S01: Acquire actual satellite orbit measurement data and preset reference orbit data.

[0024] In a preferred embodiment of this invention, the acquisition of actual satellite orbit measurement data and preset reference orbit data specifically includes: In the specific implementation process, the data acquisition step is performed first. The system collects the satellite's orbital status information in real time through the onboard global navigation satellite system receiver, star sensor, and precise orbit determination data injected by the ground control station, forming actual orbital measurement data. The actual orbital measurement data includes the current time stamp, the satellite's position and velocity vectors in the inertial frame or the Earth-centered Earth-fixed frame, and the six orbital elements calculated from the position and velocity: semi-major axis, eccentricity, orbital inclination, right ascension of the ascending node, argument of perigee, and mean perigee. Simultaneously, the system reads preset reference orbital data from the orbital parameter file stored on the satellite or injected from the ground. The reference orbital data is a strictly regressive orbital parameter under ideal conditions, also containing the timestamp, position vector, velocity vector, and six orbital elements corresponding to each sampling time, used as a comparison benchmark for the satellite's actual trajectory. To ensure temporal consistency in subsequent error assessments, the actual orbit measurement data and the reference orbit data must use the same spatiotemporal reference and be sampled and preprocessed at uniform time intervals. This includes outlier removal, smoothing filtering, and coordinate system transformation of the raw observation data to ensure that the alignment accuracy of the two types of data on the time axis meets the requirements of subsequent processing. At this point, the system has completed the initial acquisition of the satellite's actual operating status and the target orbit reference, providing complete input data for subsequent spatial error calculations.

[0025] S02: Based on the actual satellite orbit measurement data and reference orbit data, and combined with the preset time alignment strategy along the track, calculate the satellite orbit spatial error vector.

[0026] In a preferred embodiment of this invention, the step of calculating the satellite orbit spatial error vector based on the actual satellite orbit measurement data and reference orbit data, combined with a preset time alignment strategy along the track, specifically involves: After acquiring the actual orbit measurement data and the reference orbit data, the system further performs a spatial error calculation step. First, a local orbit coordinate system is established at each preset sampling point on the reference orbit. This coordinate system has its origin at the centroid of the reference satellite, with the radial axis pointing from the Earth's center to the satellite, the tangential axis lying in the orbital plane perpendicular to the radial direction and pointing in the direction of satellite motion, and the normal axis perpendicular to the orbital plane and following the right-hand rule. Together, they form a radial-tangential-normal coordinate system, denoted as the RTN coordinate system. To accurately quantify the geometric deviation of the actual orbit from the reference orbit, an RTN (radial R-tangential T-normal N) satellite orbit coordinate system is first established at sampling point k on the reference orbit. Based on the phase drift along the track direction caused by atmospheric drag on the actual orbit, a time alignment strategy along the track is introduced, i.e., at the reference sampling point... Within a nearby time window, a search algorithm is used to find the alignment moment that minimizes the tangential position deviation between the actual orbit and the reference orbit in the RTN coordinate system. Based on this alignment time Calculate the spatial error vector after eliminating the time phase difference. Its mathematical expression is: In the formula This is the rotation matrix for transforming from the Earth-centered Earth-fixed coordinate system to the RTN coordinate system. and These are the position vectors of the actual orbit and the reference orbit, respectively. Based on this, the 3D pipeline constraints are transformed into a geometric description, requiring that the magnitude of the spatial error vector always satisfy: in This is the preset pipe radius.

[0027] Specific examples Figure 2 As shown, Figure 2This paper presents an orbital spatial error assessment model based on the RTN coordinate system. The lower part of the figure represents the Earth-Centered Earth-Fixed (ECEF) coordinate system, the upper dashed line represents the preset reference orbit, and the solid line represents the actual orbit of the satellite. In traditional assessment methods, directly comparing positions at the same moment results in significant tangential errors due to phase drift along the track direction caused by atmospheric drag and other factors. This figure emphasizes the geometric meaning of the "track-time alignment strategy" introduced in this invention: instead of forcibly comparing the same time points, it searches for an optimal alignment moment on the actual orbit. This makes the actual orbital position Sampling points with reference orbit Tangential (T-axis) deviations are eliminated to the greatest extent possible. Based on this, the spatial deviation between the two is precisely quantified into a spatial error vector. It can be decomposed into radial errors used for control evaluation. and normal error This provides a reliable geometric basis for accurately determining whether a satellite has breached the given pipeline constraints.

[0028] S03: Determine whether orbit maintenance is required based on the satellite orbital spatial error vector. If orbit maintenance is required, determine the current orbital elements based on the actual orbital measurement data, and input the orbital elements into the preset deterministic reference maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase.

[0029] In a preferred embodiment of this invention, the step of determining whether orbit maintenance is required based on the satellite orbital spatial error vector, and if orbit maintenance is required, determining the current orbital elements based on the actual orbital measurement data, and inputting the orbital elements into a preset deterministic reference maintenance strategy model to obtain the theoretical tangential pulse amplitude and optimal maneuver phase, specifically: After obtaining the satellite's orbital spatial error vector, the system determines whether orbit maintenance is required. Specifically, it calculates the magnitude of the spatial error vector. If the magnitude is greater than or equal to the difference between a preset conduit radius and a safety threshold, or if it is predicted that the vector will exceed the conduit boundary in the future, the orbit maintenance process is triggered. When orbit maintenance is required, the system first extracts the satellite's current orbital elements from the actual orbital measurement data. These orbital elements include at least the semi-major axis, eccentricity vector, orbital inclination, right ascension of the ascending node, argument of perigee, and mean perigee angle. The eccentricity vector contains information on the magnitude of the eccentricity and the argument of perigee.

[0030] Subsequently, the orbital elements are input into a preset deterministic benchmark maintenance strategy model. This model first performs phase optimization, that is, it calculates the latitudinal argument used to determine the optimal ignition phase based on the difference between the frozen eccentricity vector in the reference orbital data and the eccentricity vector in the actual orbital measurement data. The calculation formula is as follows: In the formula, The eccentricity vector is frozen as a reference orbit. This represents the actual orbital eccentricity vector at the current moment. This is the obtained latitude argument. In maintaining the orbit of low-Earth orbit satellites, the latitude argument is commonly used. To indicate the satellite's position, Indicates the angle of depression at perigee. It represents the angle of approach.

[0031] After determining the phase, an objective function is constructed to maintain the peak normal error within the period close to the pipe boundary. The optimal tangential pulse amplitude is then solved using Newton's iteration method, with the iterative formula as follows: For out-of-plane control, a passive triggering strategy based on dead zone thresholds is adopted, which is applied only during tilt deviation. When the threshold is exceeded, a normal pulse is applied at the ascending or descending intersection. Make corrections.

[0032] S04: Based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, and combined with the preset two-layer solution algorithm based on Latin hypercube sampling and third-generation non-dominated sorting genetic algorithm, calculate the Pareto optimal solution set.

[0033] In a preferred embodiment of this invention, the Pareto optimal solution set is calculated based on the theoretical tangential pulse amplitude and a preset uncertainty parameter, combined with a preset two-layer solution algorithm based on Latin hypercube sampling and a third-generation non-dominated sorting genetic algorithm. Specifically: After obtaining the theoretical tangential pulse amplitude and the optimal maneuver phase, the system enters the robust optimization solution stage. First, the theoretical tangential pulse amplitude is used as the search benchmark for the decision variable, and combined with preset uncertainty parameters, they are input into a two-layer solution algorithm based on Latin hypercube sampling and a third-generation non-dominated sorting genetic algorithm to perform uncertainty propagation and multi-objective optimization. The preset uncertainty parameters include thrust amplitude error, thrust direction error, and execution timing error, which are used to simulate the random deviation of the actual thrust magnitude from the commanded thrust, the degree to which the actual thrust vector deviates from the nominal tangential axis in three-dimensional space, and the ignition position deviation caused by clock drift or command delay, respectively. Specifically, the thrust amplitude error is modeled as a random variable following a normal distribution; the thrust direction error is described by a conical distribution model, i.e., the actual thrust direction is randomly distributed within a cone centered on the nominal tangential axis with an off-axis angle of half a cone angle; and the execution timing error is modeled as a normal distribution with a mean of zero.

[0034] The preset uncertainty parameters are modeled as follows. The thrust amplitude error is used to simulate the random deviation of the actual thrust magnitude from the commanded thrust, and is modeled as follows: Secondly, a conical distribution model is adopted, such as Figure 3 As shown, the thrust direction error is described, and the off-axis angle is introduced. and azimuth To characterize the deviation of the actual thrust vector from the nominal tangential axis in three-dimensional space The degree of actual thrust direction vector In the RTN coordinate system, it is represented as: Finally, considering the maneuver phase error (execution timing error), the simulation of ignition position deviation caused by onboard clock drift or command delay is performed to determine the actual ignition phase. The model is as follows: In the formula For command phase, This represents the phase error.

[0035] Based on this, this embodiment constructs a robust multi-objective optimization model. This model uses the theoretical tangential pulse amplitude... The decision variable is the set of execution errors that follow a specific probability distribution. Given uncertain input, the optimization objective comprises two aspects: maximizing expected performance and minimizing performance volatility variance. Mathematically, this is expressed as: Among them, the first goal The first objective is to maximize the mean of the number of effective maintenance cycles, representing the system's expected performance under error disturbances; the second objective is... The goal is to minimize the standard deviation of the number of maintenance cycles, which represents the system's insensitivity to errors and its robust stability.

[0036] The two-layer solution algorithm employs a third-generation non-dominated sorting genetic algorithm for global optimization in its outer layer, and utilizes Latin hypercube sampling technology for uncertainty propagation evaluation in its inner layer. The specific execution steps are as follows: Before the outer loop begins, multiple individuals representing tangential pulse amplitudes are randomly generated within the neighborhood of the theoretical tangential pulse amplitude, forming an initial population. Each individual represents a candidate control pulse amplitude. Subsequently, each individual in the initial population is sequentially passed to the inner loop.

[0037] In the inner loop, for each incoming individual, uncertainty propagation assessment is performed. First, Latin hypercube sampling is used to perform stratified sampling of the uncertainty parameters, generating a statistically representative set of physical error samples. Specifically, the number of samplings is set to N. For each uncertainty parameter (thrust amplitude error, off-axis angle, azimuth angle, phase error), its probability distribution interval is divided into N non-overlapping sub-intervals. A sample point is randomly selected from each sub-interval, and then the parameter samples are randomly combined to form N sets of physical error samples. Compared with traditional Monte Carlo random sampling, Latin hypercube sampling can achieve higher sampling accuracy with fewer samples. The generated set of physical error samples is injected one by one into a preset high-precision orbit recursive model. This model takes the tangential pulse amplitude, optimal maneuver phase, and uncertainty parameter samples of the current individual as inputs, and recursively calculates the effective maintenance cycle number under the error disturbance. The effective maintenance cycle number is defined as the number of orbital cycles in which the magnitude of the satellite orbital spatial error vector continuously satisfies the pipe constraint from the current moment. After all N sample injections are completed, N valid maintenance cycle sample values ​​are obtained. Based on these, the statistical characteristics of the maintenance cycle count for this individual under error disturbances are calculated, namely the mean and variance of the maintenance cycle count. The mean of the maintenance cycle count reflects the expected maintenance performance of the control strategy under multiple random error disturbances, while the variance of the maintenance cycle count characterizes the sensitivity of the strategy to error disturbances; the smaller the variance, the higher the robustness and stability. The inner loop uses the evaluated mean and variance of the maintenance cycle count as the optimization objective value for this individual and passes it back to the outer loop.

[0038] After receiving the optimization objective values ​​from all individuals in the population, the outer loop performs multi-objective optimization operations. In this embodiment, maximizing the mean of the number of maintenance cycles is the first optimization objective, and minimizing the variance of the number of maintenance cycles is the second optimization objective. The third-generation non-dominated sorting genetic algorithm first performs fast non-dominated sorting on all individuals in the current population, dividing individuals into different levels according to Pareto dominance; then it calculates the reference point correlation degree and niche count for each individual to maintain population diversity; next, it generates offspring populations through tournament selection, simulated binary crossover, and polynomial mutation operators, merges the parent and offspring, and performs non-dominated sorting and selection again to form a new generation population. The above outer evolutionary process is executed iteratively until the preset maximum number of iterations is reached or the population converges to the preset termination condition. After the iteration ends, the algorithm outputs a Pareto optimal solution set, which contains a series of non-dominated solutions that achieve the optimal trade-off between expected maintenance performance and robust stability, allowing decision-makers to select the final control scheme from them according to actual task requirements.

[0039] The specific execution flow and key parameter configuration of the algorithm in this embodiment are as follows: 1) Input and Initialization: Set the outer population size N (e.g., 20), and the maximum number of iterations. (e.g., set to 10), inner layer LHS sampling number (e.g., set to 100), and randomly generate the initial population within the decision space. .

[0040] 2) Outer loop (NSGA-III multi-objective optimization): Utilizes SBX crossover and polynomial mutation operators to generate offspring populations. and merged with the parent generation to form Each individual in the population It is passed to the inner loop for evaluation.

[0041] 3) Inner loop (LHS uncertainty propagation assessment): Targeting incoming individuals. Perform the following operations: Step A (Nominal Performance Calculation): Calculate the error-free condition ( The nominal maintenance lap count under ) Step B (LHS Sampling and Injection): Generating using the inverse cumulative distribution function. Latin hypercube physical error sample set (Including amplitude, direction angle, azimuth angle, and phase error). Samples are injected one by one into the orbital recursion model to obtain the effective maintenance cycle count under different error perturbations. Step C (Calculate the objective function): Based on Based on the second sampling result, calculate the individual Expected performance target value and robust stability target value .

[0042] 4) Selection and Update (Return to Outer Layer): For the evaluated merged population... Perform fast non-dominated sorting. Perform target normalization and associate individuals with predefined reference points. Utilize niche counting mechanisms from... Screening Individuals form a new generation of population. .

[0043] 5) Output results: The iteration reached... Then, output the Pareto optimal solution set.

[0044] like Figure 4 As shown in the figure, this diagram visually illustrates the overall algorithm framework for robust orbit maintenance using a two-layer solution based on LHS-NSGA-III. The algorithm structure is clearly divided into two nested loops: the outer core is the NSGA-III multi-objective optimization module, responsible for population initialization, crossover and mutation, non-dominated sorting, and multi-objective trade-off evolution of the control variable (tangential pulse amplitude); the inner dashed box contains the LHS uncertainty propagation evaluation module, which serves as the "objective function evaluator" for the outer algorithm.

[0045] Each control strategy generated by the outer algorithm is passed to the inner layer. The inner layer uses efficient Latin hypercube stratified sampling (LHS) to map the random error parameters of the actuators (thrust amplitude, heading angle, azimuth angle, and phase error) to physical samples with equal probability, and injects them one by one into the orbital recursive calculation. This allows for the accurate calculation of the statistical response characteristics of the control strategy under random disturbances (i.e., expected maintenance performance and performance fluctuation variance). This statistical evaluation result is then returned to the outer layer to support the selection of the next generation of the population. Finally, after a set number of iterations, the algorithm outputs a Pareto optimal solution set that balances "high performance" and "high stability".

[0046] S05: Select the robust optimal pulse amplitude from the Pareto optimal solution set, and generate control commands based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, so as to realize pipeline control of the return orbit.

[0047] In a preferred embodiment of this invention, the step of selecting a robust optimal pulse amplitude from the Pareto optimal solution set and generating control commands based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, thereby achieving pipeline control of the return orbit, specifically involves: After obtaining the Pareto optimal solution set, the system selects the final robust optimal pulse amplitude from the solution set according to the current task requirements. Specifically, if the task has higher requirements for the maintenance period length, the individual with the larger expected maintenance cycle mean is selected from the solution set; if the task has higher requirements for control stability, such as ensuring no out-of-bounds errors under extreme error environments, the individual with the smaller maintenance cycle variance is selected from the solution set. In practical engineering, a compromise strategy is usually adopted, selecting the individual with the best comprehensive evaluation of mean and variance in the solution set. The evaluation function can be set as a linear weighted sum of the mean and variance with preset weights, or a balanced solution can be obtained by solving the knee point of the solution set. The system determines the tangential pulse amplitude corresponding to the selected individual as the robust optimal pulse amplitude.

[0048] Subsequently, based on the robust optimal pulse amplitude and the previously determined optimal maneuver phase, specific satellite control commands are generated. These commands include at least the magnitude of the tangential velocity increment to be applied, the latitude argument corresponding to the desired ignition time, the thrust direction, and the thrust duration. The system transmits the control commands to the satellite's attitude and orbit control computer via a space-to-ground communication link or an onboard autonomous planning module. Upon receiving the commands, the satellite actuators, when the optimal maneuver phase arrives, perform an orbital maneuver along the tangential direction with the velocity increment corresponding to the robust optimal pulse amplitude. During the maneuver, the onboard propulsion system (such as an electric thruster or a chemical thruster) outputs thrust according to the commands, correcting the semi-major axis and eccentricity. After the maneuver, the satellite continues to operate along the new orbit, and its actual trajectory will always remain within the preset pipe radius, thus achieving precise pipe control of the strict return orbit. Through the above process, the system fully utilizes the Pareto front solution of robust multi-objective optimization under the premise of explicitly quantifying the uncertainty of the actuator. This not only ensures the expected performance of long-term orbit maintenance, but also significantly reduces the risk of going out of bounds due to random errors, and ultimately achieves highly reliable and high-precision pipeline maintenance of the return orbit.

[0049] like Figure 5 As shown, in order to more intuitively verify the significant effect of the above technical solution in improving system reliability, this application provides simulation comparison verification results. Figure 5 This paper presents a comparison of the trajectory normal-radial plane (RN plane) error distribution under the same execution error (thrust amplitude, direction, and time deviation) conditions, using a conventional deterministic strategy (left figure) and the robust optimization strategy of this application (right figure). The black dashed circles in the figures represent the defined pipeline constraint boundaries. The comparison clearly shows that the error sample points of the prior art (left figure) diverge significantly and exceed the pipeline boundary, while the sample points controlled by the strategy of this application (right figure) are tightly constrained within the pipeline boundary, strongly demonstrating the advantage of this application in reducing the risk of boundary breaches.

[0050] In summary, this application first acquires actual satellite orbit measurement data and reference orbit data, and then accurately calculates the spatial error vector by combining a time alignment strategy along the track, eliminating the error assessment bias caused by neglecting phase drift in traditional methods. When it is determined that orbit maintenance is required, a deterministic benchmark strategy is used to quickly obtain the theoretical tangential pulse amplitude and optimal maneuver phase based on actual orbit elements, providing a reliable nominal benchmark for subsequent optimization. Then, using this theoretical pulse amplitude as the search center, uncertainty parameters of the actuator are introduced, and a two-layer solution algorithm based on Latin hypercube sampling and a third-generation non-dominated sorting genetic algorithm is used for uncertainty propagation and multi-objective optimization. This maximizes the expected performance of the effective maintenance orbit number while minimizing the performance fluctuation variance. Finally, the robust optimal pulse amplitude is selected from the Pareto optimal solution set to execute the orbit maneuver, enabling the control system to explicitly quantify the impact of thrust amplitude, direction, and time deviations, significantly reducing the risk of out-of-bounds movement caused by execution errors, achieving the optimal trade-off between maintenance benefits and system stability, and effectively improving the reliability and robustness of strict regression orbit pipeline control of the satellite in non-ideal environments. This application effectively solves the problem that existing technologies cannot accurately and efficiently control the pipeline for satellite return to orbit.

[0051] Example 2 Please refer to Figure 6 This is a pipeline control device for satellite return to orbit provided in the embodiments of this application.

[0052] In this embodiment, the pipeline control device for satellite return to orbit includes an acquisition module 10, a first calculation module 20, a processing module 30, a second calculation module 40, and a control module 50.

[0053] The acquisition module 10 is used to acquire actual satellite orbit measurement data and preset reference orbit data; The first calculation module 20 is used to calculate the satellite orbit spatial error vector based on the actual satellite orbit measurement data and reference orbit data, combined with a preset time alignment strategy along the track. Processing module 30 is used to determine whether orbit maintenance is required based on the satellite orbital spatial error vector. If orbit maintenance is required, it determines the current orbital elements based on the actual orbital measurement data and inputs the orbital elements into a preset deterministic reference maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase. The second calculation module 40 is used to calculate the Pareto optimal solution set based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, combined with the preset two-layer solution algorithm based on Latin hypercube sampling and the third-generation non-dominated sorting genetic algorithm. The control module 50 is used to select the robust optimal pulse amplitude from the Pareto optimal solution set and generate control commands based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, so as to realize pipeline control of the return orbit.

[0054] For ease of description and brevity, the embodiments of the device of the present invention include all the implementation methods in the above-described embodiments of the pipeline control method for satellite return to orbit, and will not be repeated here.

[0055] Example 3: This application provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the aforementioned pipeline control method for satellite reentry into orbit. The satellite return orbit pipeline control method, if implemented as a software functional unit and used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0056] Example 4 This embodiment provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements any one of the satellite return-to-orbit pipeline control methods described in Embodiment 1.

[0057] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A pipeline control method for satellite reentry orbit, characterized in that, include: Acquire actual satellite orbit measurement data and preset reference orbit data; Based on the actual satellite orbit measurement data and reference orbit data, and combined with the preset time alignment strategy along the track, the satellite orbit spatial error vector is calculated. Based on the satellite orbital spatial error vector, it is determined whether orbit maintenance is required. If orbit maintenance is required, the current orbital elements are determined based on the actual orbital measurement data, and the orbital elements are input into the preset deterministic benchmark maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase. Based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, combined with the preset two-layer solution algorithm based on Latin hypercube sampling and third-generation non-dominated sorting genetic algorithm, the Pareto optimal solution set is calculated. The robust optimal pulse amplitude is selected from the Pareto optimal solution set, and control commands are generated based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, thereby achieving pipeline control of the return orbit.

2. The pipeline control method for satellite return orbit according to claim 1, characterized in that, The satellite orbit spatial error vector is calculated based on the actual satellite orbit measurement data and reference orbit data, combined with a preset time alignment strategy along the track, specifically as follows: Establish radial, tangential, and normal orbital coordinate systems at the reference orbit sampling points; Based on the phase drift along the track direction caused by atmospheric drag on the actual orbit, within the time window of the reference orbit sampling point, a search algorithm is used to find the alignment moment that minimizes the tangential position deviation between the actual orbit and the reference orbit in the radial, tangential and normal orbit coordinate systems. Based on the alignment time, the position vector of the actual orbit is subtracted from the position vector of the reference orbit, and a coordinate transformation is performed using the rotation matrix from the preset geocentric coordinate system to the radial, tangential, and normal orbit coordinate systems to calculate the satellite orbit spatial error vector after eliminating the time phase difference.

3. The pipeline control method for satellite return orbit according to claim 1, characterized in that, The process involves determining the current orbital elements based on the actual orbital measurement data and inputting these elements into a preset deterministic benchmark maintenance strategy model to obtain the theoretical tangential pulse amplitude and optimal maneuver phase. Specifically: Based on the difference between the frozen eccentricity vector in the reference orbit data and the eccentricity vector in the actual orbit measurement data, the latitude argument used to determine the optimal ignition phase is calculated, and the latitude argument is taken as the optimal maneuver phase. With the optimization objective of maintaining the peak value of the normal error within a preset period close to the preset pipeline boundary, an objective function is constructed, and the Newton-Raphson iteration method is used to solve the objective function to obtain the theoretical tangential pulse amplitude.

4. The pipeline control method for satellite return orbit according to claim 1, characterized in that, Based on the theoretical tangential pulse amplitude and preset uncertainty parameters, and combined with a preset two-layer solution algorithm based on Latin hypercube sampling and third-generation non-dominated sorting genetic algorithm, the Pareto optimal solution set is calculated, specifically as follows: The theoretical tangential pulse amplitude and the preset uncertainty parameters are input into a two-layer solution algorithm for uncertainty propagation and multi-objective optimization. In the outer loop, the initial population of individuals with each tangential pulse amplitude is generated using the third-generation non-dominated sorting genetic algorithm. Each individual in the initial population is passed to the inner loop. In the inner loop, the uncertainty parameter is stratified and sampled using Latin hypercube sampling to generate a physical error sample set. The physical error sample set is then injected into a preset orbital recursion model to evaluate the statistical characteristics of the number of maintenance cycles of the current individual under error interference. The inner loop passes the mean and variance of the number of maintenance cycles obtained from the evaluation to the outer loop as the optimization target value for the individual. The outer loop performs non-dominated sorting and selection on all individuals according to the target value, generates a new generation of population, and iterates until the preset termination condition is met, outputting the Pareto optimal solution set.

5. The pipeline control method for satellite return orbit according to claim 4, characterized in that, The preset uncertainty parameters include thrust amplitude error, thrust direction error, and execution timing error; wherein, the thrust amplitude error is used to simulate the deviation of the actual thrust magnitude from the preset command thrust, the thrust direction error is modeled using a preset conical distribution model, and the execution timing error is used to simulate the ignition position deviation caused by clock drift or command delay.

6. A pipeline control device for satellite reentry into orbit, characterized in that, include: The acquisition module is used to acquire actual satellite orbit measurement data and preset reference orbit data; The first calculation module is used to calculate the satellite orbit spatial error vector based on the actual orbit measurement data and reference orbit data of the satellite, combined with a preset time alignment strategy along the track. The processing module is used to determine whether orbit maintenance is required based on the satellite orbital spatial error vector. If orbit maintenance is required, it determines the current orbital elements based on the actual orbital measurement data and inputs the orbital elements into a preset deterministic reference maintenance strategy model to obtain the theoretical tangential pulse amplitude and the optimal maneuver phase. The second calculation module is used to calculate the Pareto optimal solution set based on the theoretical tangential pulse amplitude and the preset uncertainty parameters, combined with the preset two-layer solution algorithm based on Latin hypercube sampling and the third-generation non-dominated sorting genetic algorithm. The control module is used to select the robust optimal pulse amplitude from the Pareto optimal solution set, and generate control commands based on the robust optimal pulse amplitude to control the satellite actuators to perform orbital maneuvers, so as to realize pipeline control of the return orbit.

7. The pipeline control device for satellite return orbit according to claim 6, characterized in that, The satellite orbit spatial error vector is calculated based on the actual satellite orbit measurement data and reference orbit data, combined with a preset time alignment strategy along the track, specifically as follows: Establish radial, tangential, and normal orbital coordinate systems at the reference orbit sampling points; Based on the phase drift along the track direction caused by atmospheric drag on the actual orbit, within the time window of the reference orbit sampling point, a search algorithm is used to find the alignment moment that minimizes the tangential position deviation between the actual orbit and the reference orbit in the radial, tangential and normal orbit coordinate systems. Based on the alignment time, the position vector of the actual orbit is subtracted from the position vector of the reference orbit, and a coordinate transformation is performed using the rotation matrix from the preset geocentric coordinate system to the radial, tangential, and normal orbit coordinate systems to calculate the satellite orbit spatial error vector after eliminating the time phase difference.

8. The pipeline control device for satellite return orbit according to claim 6, characterized in that, The process involves determining the current orbital elements based on the actual orbital measurement data and inputting these elements into a preset deterministic benchmark maintenance strategy model to obtain the theoretical tangential pulse amplitude and optimal maneuver phase. Specifically: Based on the difference between the frozen eccentricity vector in the reference orbit data and the eccentricity vector in the actual orbit measurement data, the latitude argument used to determine the optimal ignition phase is calculated, and the latitude argument is taken as the optimal maneuver phase. With the optimization objective of maintaining the peak value of the normal error within a preset period close to the preset pipeline boundary, an objective function is constructed, and the Newton-Raphson iteration method is used to solve the objective function to obtain the theoretical tangential pulse amplitude.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the pipeline control method for satellite return to orbit as described in any one of claims 1 to 5.

10. A terminal device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the pipeline control method for satellite return to orbit as described in any one of claims 1 to 5.

Images