A double closed loop physical information embedded photoelectric tracking trajectory position prediction system

Abstract

This invention provides a prediction system for photoelectric tracking trajectory positions embedded with dual closed-loop physical information, relating to the field of target trajectory tracking and prediction technology. It includes an inner-loop fuzzy prediction subsystem for performing fuzzy prediction processing on the trajectory position information set from historical time-series observation data of the target satellite to obtain a predicted trajectory position set and a prediction uncertainty set; an outer-loop orbital dynamics verification subsystem for generating a candidate predicted trajectory position set and, based on the orbital elements obtained for each candidate predicted trajectory position, performing correction processing when a candidate predicted trajectory position is deemed infeasible; and a confidence-driven control subsystem for determining the confidence level of each predicted trajectory position after correction processing, determining the target control mode based on the trend of confidence level changes, and performing control processing according to the target control mode. This solves the technical problem of control failure after a brief loss of the tracked target, achieving a stable target tracking effect.

Description

Technical Field

[0001] This invention relates to the field of target trajectory tracking and prediction technology, and in particular to a prediction system for the position of photoelectric tracking trajectory with embedded dual closed-loop physical information. Background Technology

[0002] With the rapid development of space technology, the number of satellites in orbit has increased dramatically. Precise tracking and monitoring of satellite targets is of great significance in fields such as aerospace telemetry and control, space situational awareness, and astronomical observation.

[0003] Telescopes and other photoelectric tracking devices are currently the core means of achieving high-precision tracking of satellite targets. Under ideal observation conditions, the photoelectric tracking system extracts the target position through an image processing system and generates closed-loop control commands to drive the turntable, thereby achieving continuous locking and tracking of the target.

[0004] However, in the complex field observation environment, the photoelectric tracking system is easily affected by external natural and human factors, which makes it impossible to track the target satellite effectively for a long time. There may be situations where the target satellite is temporarily lost. In such cases, the inability to quickly capture the target satellite leads to the failure of target satellite control. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a prediction system for the position of photoelectric tracking trajectory with embedded dual closed-loop physical information, so as to solve the technical problem that the target satellite cannot be smoothly tracked and controlled in the prior art.

[0006] In a first aspect, embodiments of the present invention provide a prediction system for the position of photoelectric tracking trajectory embedded with dual closed-loop physical information. The prediction system includes: an inner-loop fuzzy prediction subsystem, an outer-loop trajectory dynamics verification subsystem, and a confidence-driven control subsystem; wherein... The inner loop fuzzy prediction subsystem is used to perform fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series. The outer ring track dynamics verification subsystem is used to generate a candidate predicted trajectory position set based on the predicted trajectory position set and the prediction uncertainty set, and to calculate the number of track elements based on each candidate predicted trajectory position. When it is determined that the candidate predicted trajectory position is infeasible, a correction process is performed to obtain a corrected predicted trajectory position set. The confidence-driven control subsystem is used to determine the confidence level of each predicted trajectory position in the corrected predicted trajectory position set, and to determine a target control mode based on the changing trend of the confidence level of the predicted trajectory position, so as to control the predicted trajectory position according to the target control mode.

[0007] In conjunction with the first aspect, the present invention provides a first possible implementation of the first aspect, wherein the inner loop fuzzy prediction subsystem is further configured to preprocess the trajectory position information set and initialize a type II fuzzy rule base; the inner loop fuzzy prediction subsystem includes: a coarse-grained layer trajectory trend prediction module; The coarse-grained layer trajectory trend prediction module is used for: Based on the first time window value, extract the historical trajectory location set from the preprocessed trajectory location information set; Based on the initialized triangular membership function, and in accordance with the initialized type-two fuzzy rule base, type-two fuzzy inference processing is performed on the historical trajectory position set to obtain the predicted trajectory position set and prediction uncertainty set in the future time series.

[0008] In conjunction with the first aspect, this embodiment of the invention provides a second possible implementation of the first aspect, wherein the inner-loop fuzzy prediction subsystem further includes: a fine-grained layer residual perturbation prediction module; the fine-grained layer residual perturbation prediction module is used for: According to the second time window value, extract the target predicted trajectory location set from the predicted trajectory location set; Based on the actual observed trajectory position set corresponding to the target predicted trajectory position set, the difference between each target predicted trajectory position and the actual observed trajectory position is calculated to obtain the coarse-grained layer prediction residual set of the target predicted trajectory position set; the second time window value is smaller than the first time window value; The initial type II fuzzy rule base is updated, and based on the updated type II fuzzy rule base and the initialized Gaussian membership function, the type II fuzzy inference process is performed on the coarse-grained layer prediction residual set to obtain the fine-grained layer prediction residual set and the fine-grained layer prediction uncertainty set.

[0009] In conjunction with the first aspect, this embodiment of the invention provides a third possible implementation of the first aspect, wherein the inner loop fuzzy prediction subsystem further includes: an adaptive rule evolution module; The adaptive rule evolution module is used to perform the following processing on any target rule in the initialized type-II fuzzy rule base to achieve the update processing of the initialized type-II fuzzy rule base: Within the first time window, the number of triggers and the trigger frequency of normalization processing are counted, and the rule uncertainty corresponding to the target rule is obtained, so as to update the activity of the target rule according to the rule uncertainty; When the duration for which the activity of the target rule is less than the preset extinction threshold is greater than the duration threshold, rule extinction processing is triggered to delete the target rule from the initialized type II fuzzy rule base. When the prediction residual of the coarse-grained orbit trend prediction module is greater than the preset prediction error threshold, and / or the prediction uncertainty is greater than the preset prediction uncertainty threshold, a rule generation process is triggered to insert the newly generated rule into the initialized type II fuzzy rule base.

[0010] In conjunction with the first aspect, this embodiment of the invention provides a fourth possible implementation of the first aspect, wherein the inner loop fuzzy prediction subsystem further includes: a fusion layer uncertainty weighting module; The uncertainty weighting module of the fusion layer is used for: Using the prediction uncertainty set and the fine-grained layer prediction uncertainty set, determine the first weight corresponding to the coarse-grained layer trajectory trend prediction module and the second weight corresponding to the fine-grained layer residual disturbance prediction module; Based on the target predicted trajectory location set, the second weight, and the coarse-grained layer prediction residual set, a fused predicted trajectory location set is obtained, and based on the coarse-grained layer prediction residual set, the fine-grained layer prediction uncertainty set, the first weight, and the second weight, a fused uncertainty set is determined.

[0011] In conjunction with the first aspect, this embodiment of the invention provides a fifth possible implementation of the first aspect, wherein the fused predicted trajectory position set includes: an azimuth predicted position subset and an elevation predicted position subset; and the fused uncertainty set includes: an azimuth predicted uncertainty subset and an elevation predicted uncertainty subset. The outer ring track dynamics verification subsystem includes: a track dynamics verification module; The orbital dynamics verification module is used to receive the fused predicted trajectory position set and the fused uncertainty set, so as to perform correction processing on the candidate predicted trajectory position set, including: Based on the azimuth prediction position subset and the pitch prediction position subset, the azimuth prediction uncertainty subset and the pitch prediction uncertainty subset, and the preset confidence interval coefficient, the candidate predicted trajectory position set is generated. For each candidate trajectory position in the candidate predicted trajectory position set, the following processing is performed: The candidate trajectory position is converted from the first coordinate in the station-centered coordinate system to the second coordinate in the inertial coordinate system, and Kepler orbit fitting is performed using the second coordinate to obtain the result after Kepler orbit fitting. Based on the results of the Kepler orbit fitting process, the least squares method is used to solve for the orbital root number of the candidate trajectory position, and based on the orbital root number, the feasibility of the orbital altitude, the consistency of orbital type, and the consistency of energy conservation of the candidate trajectory position are determined. When the feasibility of orbital altitude, the consistency of orbital type, and the consistency of energy conservation all meet preset thresholds, the candidate trajectory position is determined to be physically feasible.

[0012] In conjunction with the first aspect, this embodiment of the invention provides a sixth possible implementation of the first aspect, wherein the outer ring orbit dynamics verification subsystem further includes: a prediction and correction mechanism module; The prediction and correction mechanism module is used for: When it is determined that the candidate trajectory position is not physically feasible, a set of physically feasible tracks is constructed based on the six track numbers, the distance value from the candidate trajectory position to each physically feasible track is calculated, and the physically feasible track corresponding to the minimum distance value is taken as the optimal matching track; The trajectory position on the optimal matching track that corresponds to the candidate trajectory position is taken as the corrected trajectory position of the candidate trajectory position; All the corrected trajectory positions and all physically feasible candidate trajectory positions are used as the set of predicted trajectory positions after correction.

[0013] In conjunction with the first aspect, this embodiment of the invention provides a seventh possible implementation of the first aspect, wherein the confidence level driving control subsystem includes: a confidence level division module; The confidence level classification module is used for: For each set of predicted trajectory positions after correction: Calculate the target confidence level of the predicted trajectory position after the correction process; The target confidence level is matched with a preset confidence level range to classify the target confidence level into the target confidence level range.

[0014] In conjunction with the first aspect, this invention provides an eighth possible implementation of the first aspect, wherein the confidence-driven control subsystem further includes: a control mode selection and execution module, used for: For each set of predicted trajectory positions after correction: Based on the confidence level range of the predicted trajectory position after the correction process or the changing trend of the confidence level range, the target control mode corresponding to the predicted trajectory position after the correction process is determined. According to the target control mode, the predicted trajectory position after correction is controlled.

[0015] Secondly, embodiments of the present invention provide a method for predicting the position of a photoelectric tracking trajectory by embedding dual closed-loop physical information, applied to the prediction system described in the first aspect, the prediction method comprising: The inner-loop fuzzy prediction subsystem performs fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series. The outer loop orbit dynamics verification subsystem generates a candidate predicted trajectory position set based on the predicted trajectory position set and the prediction uncertainty set, and calculates the orbital elements based on each candidate predicted trajectory position. When it is determined that the candidate predicted trajectory position is infeasible, a correction process is performed to obtain the corrected predicted trajectory position set. The confidence level of each predicted trajectory position in the corrected predicted trajectory position set is determined by the confidence level-driven control subsystem. Based on the trend of the change in the confidence level of the predicted trajectory position, a target control mode is determined so as to control the predicted trajectory position according to the target control mode.

[0016] The embodiments of the present invention bring the following beneficial effects: The present invention provides a prediction system for photoelectric tracking trajectory position embedded with dual closed-loop physical information, proposes a multi-granularity hierarchical prediction architecture, designs an uncertainty-driven adaptive rule evolution algorithm, and constructs a confidence-weighted tracking control switching logic, thereby realizing interpretable, adaptive, and highly reliable photoelectric tracking trajectory position; and the prediction system is unaffected by cloud cover, atmospheric disturbance, electromagnetic interference, etc., so as to improve the stable tracking and control of target satellites.

[0017] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the architecture of a prediction system for photoelectric tracking trajectory position embedded with dual closed-loop physical information, provided in an embodiment of the present invention. Figure 2 A schematic diagram of a target trajectory in an inertial space coordinate system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the simulation experiment dataset partitioning provided in an embodiment of the present invention; Figure 4 A schematic diagram showing the simulation and real comparison of the prediction system for embedding dual closed-loop physical information into photoelectric tracking trajectory position provided in an embodiment of the present invention; Figure 5 This is a flowchart illustrating a method for predicting the position of a photoelectric tracking trajectory by embedding dual closed-loop physical information, as provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions 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, 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] With the rapid development of space technology and the dramatic increase in the number of satellites in orbit, high-precision tracking of satellite targets using telescopes and other photoelectric tracking equipment plays a crucial role in aerospace telemetry, tracking and control, and space situational awareness. Photoelectric tracking systems typically extract the target's position in real time through an image processing system and use the miss distance (the error between the target's actual position and the current line-of-sight position of the tracking device) as feedback to form a closed-loop control system. This system drives the tracking frame to compensate for position errors, thereby achieving continuous target locking.

[0023] However, in actual field observation environments, photoelectric tracking systems are highly susceptible to interference from external environmental factors. Specifically, cloud cover can directly block or attenuate the target's light signal, leading to a sharp decrease in the target's imaging signal-to-noise ratio; atmospheric disturbances can cause beam phase distortion, resulting in jitter in the target's imaging spot; and electromagnetic interference can affect the image sensor and signal transmission link, causing increased image noise or data frame loss. Under the combined effect of these factors, image processing systems often struggle to stably extract the target's centroid, making it difficult to extract the target satellite's features, and even resulting in the loss of the target satellite's trajectory.

[0024] Existing traditional closed-loop tracking methods heavily rely on real-time miss distance as error feedback. While this control strategy can maintain high tracking accuracy when the target is stably imaged, it suffers from the following drawbacks in complex observation environments: when the target is difficult to extract or is lost, the image processing system cannot output effective miss distance information; and due to the lack of error feedback, the control loop of the traditional closed-loop tracking system is forced to be interrupted, directly leading to control failure.

[0025] Therefore, how to achieve stable tracking and control of target satellites and compensate for the inability to quickly reacquire target satellites after a brief loss has become a technical problem that urgently needs to be solved by those skilled in the art.

[0026] Based on this, the present invention provides a prediction system for the position of photoelectric tracking trajectory with embedded dual closed-loop physical information, which can achieve stable tracking and control of the target satellite based on the predicted trajectory position after the target satellite is briefly lost.

[0027] To facilitate understanding of this embodiment, a detailed description of the prediction system 100 for embedding dual closed-loop physical information into photoelectric tracking trajectory position, as disclosed in this embodiment of the invention, will be provided first.

[0028] like Figure 1 As shown, Figure 1 This is a schematic diagram of the architecture of a prediction system 100 for photoelectric tracking trajectory position embedded with dual closed-loop physical information, provided in an embodiment of the present invention. Figure 1 As shown, the prediction system 100 includes: The system consists of an inner loop fuzzy prediction subsystem 10, an outer loop orbit dynamics verification subsystem 20, and a confidence-driven control subsystem 30.

[0029] It should be noted that the system provided in this embodiment of the invention can be carried in a device or server, and the device includes, but is not limited to, photoelectric tracking devices such as telescopes.

[0030] Furthermore, the scenario applied in this embodiment is when monitoring and tracking satellites, due to interference from other influencing factors, the tracking satellite is temporarily lost. In this case, the method provided by this embodiment can quickly capture the trajectory position of the lost satellite so as to perform satellite control processing.

[0031] Among them, the inner loop fuzzy prediction subsystem 10 is used to perform fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data, so as to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series.

[0032] based on Figure 1 The inner loop fuzzy prediction subsystem 10 includes: a coarse-grained layer trajectory trend prediction module 101, a fine-grained layer residual disturbance prediction module 102, a fusion layer uncertainty weighting module 103, and an adaptive rule evolution module 104.

[0033] Specifically, the inner-loop fuzzy prediction subsystem is used to acquire the trajectory position information set of the target satellite from historical time-series observation data, and to preprocess the trajectory position information set. Optionally, the preprocessing includes, but is not limited to, normalizing the trajectory position information set. The historical time-series observation data includes the azimuth and elevation angles of the target satellite.

[0034] Meanwhile, the inner loop fuzzy prediction subsystem 10 is also used to initialize the type II fuzzy rule base, which serves as an inference container for predicting trajectory position information in future time series. It can be understood as storing predetermined trend rules as the basis for prediction.

[0035] After completing the aforementioned operations, the inner-loop fuzzy prediction subsystem 10 transmits the preprocessed trajectory position information set and the initialized type II fuzzy rule base to the coarse-grained trajectory trend prediction module 101 to trigger the coarse-grained trajectory trend prediction module 101 to extract the historical trajectory position set from the preprocessed trajectory position information set according to the first time window value; based on the initialized upper and lower triangle membership functions, and according to the initialized type II fuzzy rule base, the historical trajectory position set is subjected to type II fuzzy inference processing to obtain the predicted trajectory position set and prediction uncertainty set under the future time series.

[0036] For example, suppose the first time window value is Historical trajectory location sets are extracted from the preprocessed trajectory location information according to the first time window value. As a set of historical trajectory locations, and Represented as: ,in, This is the set of actual observation locations.

[0037] Then, based on the triangular membership function initialized in the coarse-grained layer trajectory trend prediction module 101... and ， The system initializes a type-II fuzzy rule base and performs type-II fuzzy inference to predict the trajectory and position of the target satellite in future time series. This allows for the acquisition of a predicted trajectory and position set. and prediction uncertainty . and Represented as:

[0038]

[0039] in, The output should be a real number, as per the rule. To initialize the total number of type II fuzzy rule bases.

[0040] Subsequently, the coarse-grained layer trajectory trend prediction module 101 transmits the generated predicted trajectory position set and prediction uncertainty set to the fine-grained layer residual disturbance prediction module 102.

[0041] Optionally, the fine-grained layer residual perturbation prediction module 102 is used to: extract the target predicted trajectory position set from the predicted trajectory position set according to the second time window value; calculate the difference between each target predicted trajectory position and the actual observed trajectory position based on the actual observed trajectory position set corresponding to the target predicted trajectory position set, so as to obtain the coarse-grained layer prediction residual set of the target predicted trajectory position set; the second time window value is less than the first time window value; update the initialized type II fuzzy rule base, and perform type II fuzzy inference processing on the coarse-grained layer prediction residual set based on the updated type II fuzzy rule base and the initialized Gaussian membership function, so as to obtain the fine-grained layer prediction residual set and the fine-grained layer prediction uncertainty set.

[0042] Specifically, the above steps can be processed through the following example: Assuming the second time window value is Extract the target predicted trajectory location set according to the time window from the aforementioned predicted trajectory location set, that is... This involves predicting the target's trajectory location set, which reduces computational complexity and allows for rapid acquisition of the target satellite's trajectory location. Represented as .

[0043] Furthermore, after obtaining the target predicted trajectory position set, the fine-grained layer residual perturbation prediction module 102 also needs to obtain the actual observed trajectory position corresponding to the target trajectory position set. This can be understood as needing to determine the relative difference between the actual value and the predicted value. The actual observed trajectory position can be understood as the observed value corresponding to the previous observation period. The difference between each target predicted trajectory position and the corresponding actual observed trajectory position is calculated to obtain the coarse-grained layer prediction residual set.

[0044] Assumption As a set of prediction residuals for coarse-grained layers:

[0045] in, The difference between the predicted trajectory position and the corresponding actual observed trajectory position of the target. Predict the trajectory location for the target.

[0046] This can be understood as the fine-grained layer residual perturbation prediction module 102 acquiring the error from the coarse-grained layer orbit trend prediction module 101. Based on this error, the error is adjusted, i.e., refined, and the acquired coarse-grained layer prediction residual set is input into the fine-grained layer residual perturbation prediction module 102 to achieve precise control of the error.

[0047] At this point, the prediction system 100 needs to update the parameters involved in the type II fuzzy inference processing in the coarse-grained orbit trend prediction module 101 in order to improve the accuracy of the processing. That is, the initialized triangle membership function needs to be updated. The update process is as follows: First, determine the region to be updated, i.e., the candidate region set of target trajectory locations. This can be understood as the set of target trajectory locations where, under the initial triangular membership function and the initial type-II fuzzy rule base, all matching degrees are less than a preset matching degree threshold. In other words, the target trajectory locations cannot satisfy the rule dataset. The range corresponding to these datasets is determined as the candidate region. A candidate region indicates the absence of a corresponding rule or that the membership function interval needs adjustment.

[0048] Secondly, the center of the candidate region, i.e. the data center point of the dataset that cannot satisfy the rule, is taken as the peak point of the membership function of the new rule. The peak point of the function is taken as the width of the membership function on the initial triangle, and the maximum value of the prediction uncertainty in the candidate region is taken as the width of the membership function on the initial triangle.

[0049] The update of the initial type II fuzzy rule base can be performed according to the processing steps of the subsequent adaptive rule evolution module 104.

[0050] Then, based on the updated Type II fuzzy rule base and the initialized Gaussian membership function, Type II fuzzy inference processing is performed on the coarse-grained layer prediction residual set to obtain the fine-grained layer prediction residual set and the fine-grained layer prediction uncertainty set. This allows for a better capture of the stochastic characteristics of the prediction residuals.

[0051] Understandably, following the update process of the triangular membership function, the initialized Gaussian membership function also needs to be updated. This will not be elaborated here, but it is necessary to improve the accuracy of the prediction.

[0052] For example, suppose the set of prediction residuals for the fine-grained layer is represented as The set of prediction uncertainties for fine-grained layers is represented as follows: .

[0053] in, This indicates that the uncertainty set can be determined by the difference between the true residual set and the coarse-grained layer prediction residual set. The true residual set is obtained by calculating the difference between the coarse-grained layer prediction output and the actual observed values, while the fine-grained layer prediction uncertainty set is represented as... It can be obtained in the same way as the aforementioned method for obtaining the set of prediction uncertainties.

[0054] The aforementioned update and initialization of the type II fuzzy rule base can be implemented through the adaptive rule evolution module 104 in the inner-loop fuzzy prediction subsystem. Optionally, under the first time window value, the number of triggers and the trigger frequency of the normalization process are counted, and the rule uncertainty corresponding to the target rule is obtained, so as to update the activity of the target rule according to the rule uncertainty.

[0055] It should be noted that when the inner loop fuzzy prediction subsystem preprocesses the trajectory position information set, the adaptive rule evolution module 104 synchronously triggers internal counting tools, such as counters, to accumulate the number of triggers and frequency of triggers involving normalization processing during the preprocessing process, as well as the target rules used by the coarse-grained layer trajectory trend prediction module 101, and determines the activity level of the target rules.

[0056] Assume the rule structure corresponding to the target rule r is as follows: Activity level can be expressed as: ,in, This indicates the normalized trigger frequency. This represents the number of times the event will be triggered. This represents the uncertainty of the output corresponding to the rule.

[0057] Furthermore, if the duration for which the activity level of the target rule is less than the preset extinction threshold is greater than the duration threshold, then rule extinction processing is triggered to delete the target rule from the initialized Type II fuzzy rule base.

[0058] At this point, the target rule is updated for activity within the first time window, and the activity is monitored simultaneously. If the duration of the activity being less than the preset extinction threshold is greater than the duration threshold, it indicates that the current rule is unavailable and the target rule needs to be deleted.

[0059] The preset extinction threshold and duration threshold are set by those skilled in the art based on actual circumstances. Since the thresholds set vary depending on the situation, they are not specified here.

[0060] When the prediction residual of the coarse-grained layer trajectory trend prediction module 101 is greater than the preset prediction error threshold, and / or the prediction uncertainty is greater than the preset prediction uncertainty threshold, the rule generation process is triggered to insert the newly generated rule into the initialized type II fuzzy rule base.

[0061] This can be understood as follows: when the error and uncertainty of the prediction increase, it indicates that the accuracy of the prediction is insufficient, and new rules should be added to ensure the accuracy of the prediction.

[0062] Correspondingly, the initial process for the new rules is as follows: load the updated triangle membership function and the updated Gaussian triangle membership function to obtain the boundary range corresponding to the new rules; then obtain the mean value of the trajectory position data within the new boundary range, and use this mean value as the initial output of the rule base, that is, perform loading processing on the rule base; finally, insert the new rules into the initialization of the type II fuzzy rule base and recalculate the coverage relationship between the rules.

[0063] After updating the initial type II fuzzy rule base, the adaptive rule evolution module 104 transmits the updated type II fuzzy rule base to the coarse-grained layer trajectory trend prediction module 101 and the fine-grained layer residual perturbation prediction module 102 for subsequent processing.

[0064] It is conceivable that this operation is not just a single update, but a dynamic process of updating and providing feedback.

[0065] After obtaining the fine-grained layer prediction residual set and the fine-grained layer prediction uncertainty set, the fusion layer uncertainty weighting module in the inner loop fuzzy prediction subsystem will be triggered to calculate the fusion weights in order to determine the fused prediction trajectory position set and the fused uncertainty set.

[0066] Optionally, the fusion layer uncertainty weighting module uses the prediction uncertainty set and the fine-grained layer prediction uncertainty set to determine the first weight corresponding to the coarse-grained layer trajectory trend prediction module 101 and the second weight corresponding to the fine-grained layer residual disturbance prediction module; based on the target predicted trajectory position set, the second weight, and the coarse-grained layer prediction residual set, it obtains the fused predicted trajectory position set, and based on the coarse-grained layer prediction residual set, the fine-grained layer prediction uncertainty set, the first weight, and the second weight, it determines the fused uncertainty set.

[0067] Based on the previous example, perform the above processing as follows: The first weight corresponding to the coarse-grained layer orbit trend prediction module 101 is calculated using the following formula:

[0068] in, As the first weight, To predict the uncertainty set, This represents the set of prediction uncertainties for the fine-grained layer.

[0069] The second weight corresponding to the fine-grained layer residual perturbation prediction module is calculated using the following formula:

[0070] in, The second weight is used, and the other parameters are the same as those mentioned above, so they will not be repeated here.

[0071] The fused predicted trajectory location set can be obtained using the following formula:

[0072] in, This is the fused predicted trajectory location set. Predict the trajectory location set for the target. Second weight and Coarse-grained layer predicts residual sets.

[0073] The fused uncertainty set can be obtained using the following formula:

[0074] in, This is the fused uncertainty set. The other parameters are the same as described above and will not be repeated here.

[0075] In summary, the inner ring fuzzy prediction subsystem achieves rapid prediction capability for target satellites that are temporarily lost. Next, the prediction system 100 provided in this embodiment of the invention will trigger the outer ring orbit dynamics verification subsystem 20 to perform dynamic verification processing on the prediction.

[0076] Optionally, there is an outer ring track dynamics verification subsystem 20, which includes a track dynamics verification module 201 and a prediction and correction mechanism module 202.

[0077] The orbital dynamics verification module receives the fused predicted trajectory position set and the fused uncertainty set to perform correction processing on the candidate predicted trajectory position set, including: Based on subsets of predicted azimuth and pitch positions, subsets of predicted azimuth and pitch uncertainties, and preset confidence interval coefficients, a set of candidate predicted trajectory positions is generated. For each candidate trajectory position in the set, the following processing is performed: the candidate trajectory position is converted from the first coordinate in the station-centered coordinate system to the second coordinate in the inertial coordinate system, and Keplerian orbit fitting is performed using the second coordinate to obtain the result after Keplerian orbit fitting. Based on the obtained result after Keplerian orbit fitting, the least squares method is used to solve for the orbital root number of the candidate trajectory position, and based on the orbital root number, the feasibility of the orbital altitude, the consistency of orbital type, and the consistency of energy conservation of the candidate trajectory position are determined. When the feasibility of orbital altitude, the consistency of orbital type, and the consistency of energy conservation all meet the preset thresholds, the candidate trajectory position is determined to be physically feasible.

[0078] For example, the steps for converting the first coordinate in the station-centered coordinate system to the second coordinate in the inertial coordinate system are as follows: Assume the satellite data obtained by the observation station is in polar coordinates in the station-centric plane coordinate system. r , , ),in, r The distance from the satellite to the observation station. The azimuth angle of the satellite to the observation station. This is the elevation angle from the satellite to the observation station. Convert this to Cartesian coordinates in the station's tangent plane coordinate system. x , y , z )for:

[0079] Given that the geodetic coordinates of station P are ( L,B , H ),but P The coordinates of the point in the WGS84 coordinate system ( X 0, Y 0, Z 0 for:

[0080] In the above formula, N is the radius of curvature of the zonal loop. ; e For the eccentricity of the ellipse, .

[0081] The coordinates of the station center in the WGS84 coordinate system are known to be: P ( X 0, Y 0, Z 0 .satellite Q Rectangular coordinates of the point in the tangential plane coordinate system Q ( x , y , z Convert to WGS84 coordinates ( X , Y , Z ;

[0082] The coordinate transformation during the observation process needs to take into account the Earth's rotation; Greenwich Mean Time (GMT) transformation matrix. for:

[0083] Therefore, the second coordinates in the inertial coordinate system are finally obtained. for:

[0084] Select a segment of trajectory data observed at a certain station over a period of time, where the station's parameters are as follows.

[0085]

[0086] Figure 2 The figure shows a schematic diagram of a target trajectory in an inertial space coordinate system provided by an embodiment of the present invention. As shown in the figure, the actual target space trajectory and the fitted target space trajectory are basically consistent.

[0087] It should be noted that the fused predicted trajectory position set includes: the azimuth predicted position subset and the pitch predicted position subset; the fused uncertainty set includes: the azimuth predicted uncertainty subset and the pitch predicted uncertainty subset.

[0088] When performing dynamic physics verification, it is necessary to process the system from the perspective of the horizontal axis (A-axis) and the pitch axis (E-axis), that is, to process the fused predicted trajectory position into two parts.

[0089] For the trajectory position sets corresponding to these two angles, generate candidate predicted trajectory position sets. .

[0090]

[0091] in, Represented as a subset of predicted azimuth positions. This is represented as a subset of azimuth prediction uncertainty. Represented as a subset of pitch angle predicted positions, Represented as a subset of pitch angle prediction uncertainty; k The confidence interval coefficients are preset and can be selected. k =2.

[0092] For each candidate predicted trajectory location set, the following processing is performed: The candidate trajectory position is converted from the first coordinate in the station-centered coordinate system to the second coordinate in the inertial coordinate system. The second coordinate is then used for Kepler orbit fitting to obtain the result after Kepler orbit fitting.

[0093] Based on the results of Kepler orbit fitting, the least squares method is used to solve for the six orbital roots of the candidate trajectory positions. ,in For the semi-major axis of the track, For eccentricity, For the track inclination angle, Right ascension of the ascending node, The perigee argument, It is the true near point angle.

[0094] Based on the six numbers obtained, the following physical constraints are determined: Feasibility of the orbital altitude: . The average radius of the Earth These are the upper and lower bound thresholds. Orbit type consistency: The resulting orbit type should be low Earth orbit, medium Earth orbit, or geostationary orbit. Energy conservation principle: the orbital energy of a celestial body is greater than its mechanical energy. It should meet the required scope.

[0095] When all the aforementioned physical constraints meet the preset threshold, it can be determined that the candidate trajectory location is physically feasible.

[0096] The preset threshold can be the threshold for the feasibility of the orbital altitude: The threshold for orbit type consistency is T, and T must fall within the period range of the corresponding orbit type (LEO: 88~120 min; MEO: 2~24 h; GEO: 1436±30 min); the threshold for energy conservation consistency is the relative deviation of specific mechanical energy. It shall not exceed 5%; and shall be set by those skilled in the art.

[0097] Correspondingly, if the candidate predicted trajectory position does not meet any of the above physical constraints, then the candidate predicted trajectory position needs to be corrected.

[0098] At this point, it is necessary to call Figure 1 The prediction correction mechanism module 202 in the outer ring track dynamics verification subsystem 20 is used to: construct a set of physically feasible tracks based on the six track roots when it is determined that the candidate trajectory position is not physically feasible; calculate the distance value from the candidate trajectory position to each physically feasible track; and take the physically feasible track corresponding to the minimum distance value as the optimal matching track; take the trajectory position on the optimal matching track corresponding to the candidate trajectory position as the corrected trajectory position of the candidate trajectory position; and take all the corrected trajectory positions and all the physically feasible candidate trajectory positions as the set of predicted trajectory positions after correction.

[0099] For example, the set of physically feasible orbits can be constructed based on the six orbital elements as follows: (1) The six orbital elements obtained by least-squares fitting of candidate trajectory positions Centered on the orbital altitude, the preset thresholds corresponding to the orbital type consistency and energy conservation consistency are used as boundaries to determine the feasible domain of the six orbital roots. (2) Within the feasible region, with Sampling is performed around the center to obtain multiple sets of orbital six-root numbers; (3) Substitute the six roots of each orbit into the Kepler orbit equation to generate the corresponding orbit, and calculate the fitting residual of each orbit to the candidate orbit position based on the candidate trajectory position used for Kepler orbit fitting. Retain the orbits with fitting residuals lower than the preset threshold to form a set of physically feasible orbits.

[0100] In summary, the outer loop orbit dynamics verification subsystem completes the physical verification process for the predictions. This allows for confidence level division based on the corrected predicted trajectory position set, enabling the execution of corresponding control actions according to the confidence level.

[0101] First, the prediction system 100 triggers the confidence level division module 301 in the confidence level drive control subsystem 30. This allows the confidence level division module to calculate the target confidence level for each corrected predicted trajectory position in the set of corrected predicted trajectory positions. The target confidence level is then matched with a preset confidence level interval to classify the target confidence level into that interval. The target confidence level of the trajectory position is the fusion uncertainty. .

[0102] Based on the preset confidence level interval in the confidence level division module, the calculated target confidence level is matched with the confidence level interval to determine the target confidence level interval of the predicted trajectory position after the current correction.

[0103] Confidence level range The determination rule is as follows:

[0104] Correspondingly, a confidence level greater than 0.9 corresponds to an extremely high confidence level (VH), indicating that the corrected predicted trajectory position is observable; a confidence level between 0.6 and 0.9 corresponds to a high confidence level (H), indicating that the corrected predicted trajectory position is highly reliable and basically observable; a confidence level between 0.5 and 0.6 corresponds to a medium confidence level (M), indicating that the corrected predicted trajectory position is reliable, but the target is partially occluded, the prediction is uncertain, the target is lost, or there is strong perturbation; and a confidence level less than 0.5 corresponds to four low confidence levels (L), indicating that the corrected predicted trajectory position is unreliable.

[0105] Then, the prediction system 100 triggers the control mode selection and execution module 302 in the confidence level driven control subsystem; so that the control mode selection and execution module determines the target control mode corresponding to the predicted trajectory position after correction for each predicted trajectory position in the set of corrected predicted trajectory positions: based on the confidence level interval or the changing trend of the confidence level interval of the corrected predicted trajectory position; and performs control processing on the corrected predicted trajectory position according to the target control mode.

[0106] Once the target confidence level range of the predicted trajectory position after correction is determined, the corresponding control mode is determined based on the current target confidence level range or whether the target confidence level range has changed.

[0107] Control mode selection: Confidence level The rating changes from VH to H, and execution mode one is adopted; confidence level The rating changes from H to M, and mode two is executed; confidence level The level is L, and the execution mode is three; the confidence level is changed from L to M, and the execution mode is restored to four.

[0108] Correspondingly, Mode 1 (Normal Closed-Loop Tracking): Standard PID closed-loop control; Mode 2 (Predictive Guided Tracking): Weighted average of closed-loop control quantity and predicted control quantity; Mode 3 (Dynamic Prior): Reduce prediction weight and increase trajectory dynamics prior weight; Mode 4 (Smooth Transition Recapture): Weighted average of control quantity before and after recapture.

[0109] In summary, the trajectory tracking of a briefly lost target satellite is completed. A multi-granularity hierarchical prediction architecture decouples the orbital trend from short-term disturbances and adaptively fuses them based on uncertainty, constructing an adaptive rule evolution mechanism that provides high-precision prediction output and uncertainty estimation. Simultaneously, the outer-loop orbital dynamics verification applies physical feasibility constraints and corrections to the inner-loop prediction output, forming a prediction-correction dual closed loop to ensure physical consistency of the prediction results. Finally, the prediction uncertainty is converted into a confidence level, driving the gradual switching of four control modes to achieve prediction-control closed-loop feedback.

[0110] Based on this, it is possible to provide physically feasible, high-confidence adaptive trajectory position prediction for the photoelectric tracking system during the period when the target satellite is lost, guide the tracking equipment to operate smoothly, and thus achieve smooth tracking and control of the target satellite.

[0111] To visually demonstrate the performance of the prediction system provided in this embodiment of the invention, the following simulation experiment was conducted. The simulation experiment was conducted in a gondola telescope system, and the data used was observation data from a certain observation station at a certain time.

[0112] Specifically, the telescope system's spatial and orbital motion control during tracking is divided into two directions: the A-axis and the E-axis. The A-axis controls horizontal azimuth movement, and the E-axis controls elevation movement. A coordinate system is established with the observation station as the center. The observed data is in polar coordinates within the Earth's tangent plane coordinate system, which is then converted into a planar coordinate system centered on the Earth's center. Through coordinate system transformation, the satellite's polar coordinate position trajectory can be converted into azimuth and elevation angle position data from the satellite to the observation station. Using a set of position target guidance data obtained from a test range, the guidance position data for the A-axis and E-axis are obtained as follows: Figure 3 As shown, Figure 3 This is a schematic diagram illustrating the partitioning of the simulation experiment dataset provided in this embodiment of the invention. The first 300 data points are used as the training set, and the last 243 data points are used as the test set.

[0113] Subsequently, the prediction results for the training and test sets are as follows: Figure 4 As shown, Figure 4 This is a schematic diagram comparing the simulation and real-world performance of a prediction system for embedding dual-closed-loop physical information into photoelectric tracking trajectory positions, as provided in an embodiment of the present invention. Figure 4 As shown, Figure 4 (a) shows the trajectory prediction results for the training set. Figure 4(b) shows the trajectory prediction results for the test set; from Figure 4 It can be seen that the prediction error of the A-axis trajectory data on the training set is 0.001238, and the prediction error on the test set is 0.001651. The prediction error of the E-axis trajectory data on the training set is 0.001334, and the prediction error on the test set is 0.001703.

[0114] To simulate a target loss scenario and evaluate the effectiveness of the invention, a simulation experiment was conducted. The simulation experiment was run in Matlab 2023b, and the data consisted of the A-axis and E-axis training sets. The Kalman Filter is the classic Kalman prediction filter method, and the LSTM is the classic Long Short-Term Memory network. The simulation results are shown in Table 1, which is the simulation experiment data table.

[0115] Table 1 Simulation Experiment Data Table

[0116] As shown in Table 1, (1) the method of the present invention is significantly better than the comparative method in all loss duration intervals. When the loss duration is <5s, the recapture rate of the present invention is 96.4%, which is 18.1 percentage points higher than the 78.3% of the Kalman Filter and 20.0 percentage points higher than the 76.4% of the LSTM. When the loss duration is 5-15s, the recapture rate of the present invention is 87.2%, which is 34.8 and 33.1 percentage points higher than the Kalman Filter and LSTM, respectively. When the loss duration is >15s, the recapture rate of the present invention is 68.7%, which is 40.1 and 37.5 percentage points higher than the Kalman Filter and LSTM, respectively.

[0117] (2) As the loss duration increases, the recapture rate of all three methods shows a decreasing trend, but the decrease is the smallest for the method of this invention. The recapture rate of this invention decreased from 96.4% to 68.7%, a decrease of 27.7 percentage points; that of Kalman Filter decreased from 78.3% to 28.6%, a decrease of 49.7 percentage points; and that of LSTM decreased from 76.4% to 31.2%, a decrease of 45.2 percentage points. This indicates that the method of this invention has stronger robustness to long-term loss scenarios.

[0118] (3) In terms of overall success rate, the recapture rate of the present invention reaches 84.2%, which is 31.1 and 30.3 percentage points higher than that of Kalman Filter (53.1%) and LSTM (53.9%), respectively, with an improvement of more than 58%, indicating that the present invention has significant advantages in target loss and recapture scenarios.

[0119] Figure 5 This is a flowchart illustrating a method for predicting the position of a photoelectric tracking trajectory by embedding dual closed-loop physical information, as provided in an embodiment of the present invention. Figure 5As shown, the prediction method is applied to, for example Figure 1 The prediction system shown. Prediction methods include: S10. Through the inner loop fuzzy prediction subsystem, perform fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series using the prediction uncertainty set and the fine-grained layer prediction uncertainty set. S20. Through the outer ring track dynamics verification subsystem, based on the predicted trajectory position set and the candidate predicted trajectory position set generated by the predicted uncertainty set and the fine-grained layer predicted uncertainty set, and based on the number of track elements obtained for each candidate predicted trajectory position, when it is determined that it is not feasible to use the candidate predicted trajectory position set and the fine-grained layer predicted uncertainty set, a correction process is performed to obtain the corrected predicted trajectory position set. S30. Through the confidence-driven control subsystem, determine the confidence level of each predicted trajectory position in the predicted trajectory position set after correction using the predicted uncertainty set and the fine-grained layer predicted uncertainty set, and determine the target control mode based on the trend of the change of the confidence level of the predicted trajectory position using the predicted uncertainty set and the fine-grained layer predicted uncertainty set, so as to control the predicted trajectory position using the predicted uncertainty set and the fine-grained layer predicted uncertainty set according to the target control mode.

[0120] The prediction method provided in this application embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0121] The flowcharts and block diagrams in the accompanying drawings illustrate systems and methods according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0122] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0123] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0124] The foregoing has described specific embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0125] In the description of the embodiments of this application, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the embodiments of this application. In the embodiments of this application, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in the embodiments of this application, as well as the features of different embodiments or examples.

[0126] Furthermore, 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 at least one of that feature. In the description of embodiments of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0127] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order according to the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0128] It should be noted that the terminals involved in the embodiments of this application may include, but are not limited to, personal computers (PCs), personal digital assistants (PDAs), wireless handheld devices, tablet computers, mobile phones, MP3 players, MP4 players, etc.

[0129] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0130] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0131] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0132] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present application should be included within the scope of protection of the present application.

Claims

1. A prediction system for photoelectric tracking trajectory position embedded with dual closed-loop physical information, characterized in that, The prediction system includes: an inner-loop fuzzy prediction subsystem, an outer-loop orbit dynamics verification subsystem, and a confidence-driven control subsystem; wherein... The inner loop fuzzy prediction subsystem is used to perform fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series. The outer ring track dynamics verification subsystem is used to generate a candidate predicted trajectory position set based on the predicted trajectory position set and the prediction uncertainty set, and to calculate the number of track elements based on each candidate predicted trajectory position. When it is determined that the candidate predicted trajectory position is infeasible, a correction process is performed to obtain a corrected predicted trajectory position set. The confidence-driven control subsystem is used to determine the confidence level of each predicted trajectory position in the corrected predicted trajectory position set, and to determine a target control mode based on the changing trend of the confidence level of the predicted trajectory position, so as to control the predicted trajectory position according to the target control mode.

2. The prediction system according to claim 1, characterized in that, The inner loop fuzzy prediction subsystem is also used to preprocess the trajectory position information set and initialize the type II fuzzy rule base; the inner loop fuzzy prediction subsystem includes: a coarse-grained layer trajectory trend prediction module; The coarse-grained layer trajectory trend prediction module is used for: Based on the first time window value, extract the historical trajectory location set from the preprocessed trajectory location information set; Based on the initialized triangular membership function, and in accordance with the initialized type-two fuzzy rule base, type-two fuzzy inference processing is performed on the historical trajectory position set to obtain the predicted trajectory position set and prediction uncertainty set in the future time series.

3. The prediction system according to claim 2, characterized in that, The inner-loop fuzzy prediction subsystem further includes: a fine-grained layer residual perturbation prediction module; the fine-grained layer residual perturbation prediction module is used for: According to the second time window value, extract the target predicted trajectory location set from the predicted trajectory location set; Based on the actual observed trajectory position set corresponding to the target predicted trajectory position set, the difference between each target predicted trajectory position and the actual observed trajectory position is calculated to obtain the coarse-grained layer prediction residual set of the target predicted trajectory position set; the second time window value is smaller than the first time window value; The initial type II fuzzy rule base is updated, and based on the updated type II fuzzy rule base and the initialized Gaussian membership function, the type II fuzzy inference process is performed on the coarse-grained layer prediction residual set to obtain the fine-grained layer prediction residual set and the fine-grained layer prediction uncertainty set.

4. The prediction system according to claim 2, characterized in that, The inner loop fuzzy prediction subsystem also includes: an adaptive rule evolution module; The adaptive rule evolution module is used to perform the following processing on any target rule in the initialized type-II fuzzy rule base to achieve the update processing of the initialized type-II fuzzy rule base: Within the first time window, the number of triggers and the trigger frequency of normalization processing are counted, and the rule uncertainty corresponding to the target rule is obtained, so as to update the activity of the target rule according to the rule uncertainty; When the duration for which the activity of the target rule is less than the preset extinction threshold is greater than the duration threshold, rule extinction processing is triggered to delete the target rule from the initialized type II fuzzy rule base. When the prediction residual of the coarse-grained orbit trend prediction module is greater than the preset prediction error threshold, and / or the prediction uncertainty is greater than the preset prediction uncertainty threshold, a rule generation process is triggered to insert the newly generated rule into the initialized type II fuzzy rule base.

5. The prediction system according to claim 3, characterized in that, The inner loop fuzzy prediction subsystem also includes: a fusion layer uncertainty weighting module; The uncertainty weighting module of the fusion layer is used for: Using the prediction uncertainty set and the fine-grained layer prediction uncertainty set, determine the first weight corresponding to the coarse-grained layer trajectory trend prediction module and the second weight corresponding to the fine-grained layer residual disturbance prediction module; Based on the target predicted trajectory location set, the second weight, and the coarse-grained layer prediction residual set, a fused predicted trajectory location set is obtained, and based on the coarse-grained layer prediction residual set, the fine-grained layer prediction uncertainty set, the first weight, and the second weight, a fused uncertainty set is determined.

6. The prediction system according to claim 5, characterized in that, The fused predicted trajectory position set includes: an azimuth predicted position subset and an elevation predicted position subset; the fused uncertainty set includes: an azimuth predicted uncertainty subset and an elevation predicted uncertainty subset; The outer ring track dynamics verification subsystem includes: a track dynamics verification module; The orbital dynamics verification module is used to receive the fused predicted trajectory position set and the fused uncertainty set, so as to perform correction processing on the candidate predicted trajectory position set, including: Based on the azimuth prediction position subset and the pitch prediction position subset, the azimuth prediction uncertainty subset and the pitch prediction uncertainty subset, and the preset confidence interval coefficient, the candidate predicted trajectory position set is generated. For each candidate trajectory position in the candidate predicted trajectory position set, the following processing is performed: The candidate trajectory position is converted from the first coordinate in the station-centered coordinate system to the second coordinate in the inertial coordinate system, and the Kepler orbit fitting process is performed using the second coordinate to obtain the result after Kepler orbit fitting. Based on the results of the Kepler orbit fitting process, the least squares method is used to solve for the orbital root number of the candidate trajectory position, and based on the orbital root number, the feasibility of the orbital altitude, the consistency of orbital type, and the consistency of energy conservation of the candidate trajectory position are determined. When the feasibility of orbital altitude, the consistency of orbital type, and the consistency of energy conservation all meet preset thresholds, the candidate trajectory position is determined to be physically feasible.

7. The prediction system according to claim 6, characterized in that, The outer ring track dynamics verification subsystem also includes: a prediction and correction mechanism module; The prediction and correction mechanism module is used for: When it is determined that the candidate trajectory position is not physically feasible, a set of physically feasible tracks is constructed based on the six track numbers, the distance value from the candidate trajectory position to each physically feasible track is calculated, and the physically feasible track corresponding to the minimum distance value is taken as the optimal matching track; The trajectory position on the optimal matching track that corresponds to the candidate trajectory position is taken as the corrected trajectory position of the candidate trajectory position; All the corrected trajectory positions and all physically feasible candidate trajectory positions are used as the set of predicted trajectory positions after correction.

8. The prediction system according to claim 1, characterized in that, The confidence level driven control subsystem includes: a confidence level classification module; The confidence level classification module is used for: For each set of predicted trajectory positions after correction: Calculate the target confidence level of the predicted trajectory position after the correction process; The target confidence level is matched with a preset confidence level range to classify the target confidence level into the target confidence level range.

9. The prediction system according to claim 1, characterized in that, The confidence-driven control subsystem further includes: a control mode selection and execution module, used for: For each set of predicted trajectory positions after correction: Based on the confidence level range of the predicted trajectory position after the correction process or the changing trend of the confidence level range, the target control mode corresponding to the predicted trajectory position after the correction process is determined. According to the target control mode, the predicted trajectory position after correction is controlled.

10. A method for predicting the position of a photoelectric tracking trajectory by embedding dual closed-loop physical information, characterized in that, The prediction method is applied to the prediction system as described in any one of claims 1-9, and the prediction method includes: The inner-loop fuzzy prediction subsystem performs fuzzy prediction processing on the trajectory position information set of the target satellite in the historical time series observation data to obtain the predicted trajectory position set and prediction uncertainty set of the target satellite in the future time series. The outer loop orbit dynamics verification subsystem generates a candidate predicted trajectory position set based on the predicted trajectory position set and the prediction uncertainty set, and calculates the orbital elements based on each candidate predicted trajectory position. When it is determined that the candidate predicted trajectory position is infeasible, a correction process is performed to obtain the corrected predicted trajectory position set. The confidence level of each predicted trajectory position in the corrected predicted trajectory position set is determined by the confidence level-driven control subsystem. Based on the trend of the change in the confidence level of the predicted trajectory position, a target control mode is determined so as to control the predicted trajectory position according to the target control mode.

Images