A sar satellite baseline prediction analysis method and system based on tle data

Abstract

This invention provides a SAR satellite baseline prediction and analysis method and system based on TLE data, applicable to the fields of SAR satellite orbit analysis and repeat orbit mission planning. The method uses the SAR satellite TLE orbital elements as input, relies on the SGP4 module to derive sub-satellite point time-series data and completes standardization verification; it locates the ascending node interval of each orbit by scanning latitude zero points, and simultaneously corrects for longitude jumps across ±180° meridians during interpolation. It calculates the longitude closure error based on the first ascending node and performs value domain normalization, combining the Earth's equatorial radius to calculate a meter-level two-dimensional equivalent baseline. It employs local trough identification combined with a 2000m threshold to filter short baseline repeat orbit windows, and finally integrates all analysis data to output structured indicators and orbital baseline change curves. The entire process automates satellite repeat orbit baseline prediction, efficiently supporting SAR interferometric observation mission planning.

Description

Technical Field

[0001] This invention relates to the field of SAR satellite orbit analysis and repeat orbit mission planning technology, and in particular to a SAR satellite baseline prediction analysis method and system based on TLE data. Background Technology

[0002] SAR satellites possess all-weather, 24 / 7 observation capabilities, and have significant application value in fields such as deformation monitoring, disaster emergency response, marine monitoring, resource surveys, and military reconnaissance. In scenarios involving repeated orbit observations and short baseline data screening, the degree of plane closure between satellite orbits at different times directly affects subsequent image selection and mission scheduling.

[0003] TLE orbital element data is characterized by its ease of acquisition, frequent updates, and suitability for rapid extrapolation. Using the existing SGP4 orbital extrapolation module, satellite nadir time-series data within a preset time range can be obtained from TLE data, thus providing fundamental data for multiple orbit prediction and baseline analysis. However, the large number of continuously sampled data points and the fact that trajectories often cross ±180° meridians make manual interpretation or simple differencing at the original sampling points prone to inefficiency, abrupt longitude jumps, and misidentification of false multiple orbit points.

[0004] Therefore, a method is needed that can automatically extract the ascending node, correct longitude loops, calculate two-dimensional plane baselines, and filter short baseline heavy rail windows using TLE / SGP4 extrapolation results.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of this application, a SAR satellite baseline prediction and analysis method based on TLE data is provided, comprising: acquiring SAR satellite TLE orbital element data, calling the SGP4 orbit extrapolation module to complete orbit extrapolation, and generating a nadir point time series dataset containing time, latitude, longitude, and altitude; performing data column identification on the dataset, completing time standardization and data validity verification, and removing abnormal and null value sampled data; traversing the verified nadir point time series data, identifying the zero-crossing intervals where latitude transitions from negative to positive values, and locking the ascending node interval corresponding to each orbit; performing linear interpolation operations on each ascending node interval, simultaneously performing longitude orbital correction during the interpolation process, and solving for the time, longitude, and altitude parameters corresponding to the ascending node of each orbital cycle; and using the longitude of the first ascending node interval as the basis for the prediction and analysis. Using degrees as a reference value, the longitude closure error of the remaining ascending nodes is calculated for each orbit, and the error range is normalized to the continuous interval [-180°, 180°]. Combined with the Earth's equatorial radius parameter, the normalized ascending node longitude closure error is converted into a meter-level two-dimensional sub-satellite equivalent plane baseline. Local troughs are identified in the baseline numerical sequence of all orbital orbits, and candidate re-orbit orbits corresponding to local minimum baseline values ​​are selected. A fixed threshold of 2000m is set to filter the candidate re-orbit orbits, retaining samples with baseline values ​​not exceeding the threshold, thus obtaining an effective short baseline re-orbit window. The equivalent plane baseline calculation results and the selected short baseline re-orbit window information are integrated to uniformly output the re-orbit time, re-orbit period, baseline value, and orbital altitude data, and simultaneously generate an orbital orbit-baseline change analysis curve.

[0008] Another aspect of this application is a SAR satellite baseline prediction and analysis system based on TLE data, used to execute executable instructions to perform the above-described SAR satellite baseline prediction and analysis method based on TLE data.

[0009] According to another aspect of this application, an electronic device includes: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-described SAR satellite baseline prediction analysis method based on TLE data by executing the executable instructions.

[0010] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described SAR satellite baseline prediction analysis method based on TLE data.

[0011] The beneficial technical effects of this application are as follows: (1) Using publicly available TLE orbital elements as the data source, there is no need to rely on precise orbital ephemeris, making data acquisition convenient and cost-effective; (2) The time series data of the sub-satellite point is automatically generated by the SGP4 model. Combined with standardization processing such as latitude zero-crossing identification, longitude circumference correction, and linear interpolation, the parameters of the ascending node are accurately extracted. (3) Calculate the longitude closure error based on the ascending node of the first orbit and normalize it to the range of [-180°, 180°]. Combine the Earth's equatorial radius with the conversion to a meter-level two-dimensional equivalent plane baseline to quantitatively characterize the degree of orbital plane closure. (4) Local valley identification combined with a 2000m fixed threshold is used to filter short baseline re-track windows, and re-track prediction is completed automatically, which greatly improves the efficiency of task planning. (5) Output structured data reports and visualization graphs to intuitively show the baseline change trend, which can directly support the scheduling of SAR interferometric observation tasks, screening of short baseline images and quantitative analysis of orbit closure deviation.

[0012] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0013] Figure 1 A flowchart of a SAR satellite baseline prediction analysis method based on TLE data provided in an embodiment of this application is shown; Figure 2 This is a diagram from a NORADID51284 satellite reorbit analysis report; Figure 3 This is a diagram from a NORADID51822 satellite reorbit analysis report; Figure 4 This is a diagram from a NORADID32376 satellite reorbit analysis report; Figure 5 This is a diagram from a NORADID37216 satellite reorbit analysis report; Figure 6 This is a diagram from a NORADID39634 satellite reorbit analysis report. Detailed Implementation

[0014] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0015] In one implementation, Figure 1 A schematic flowchart of a SAR satellite baseline prediction analysis method based on TLE data according to an embodiment of this application is shown.

[0016] S101: Obtain SAR satellite TLE orbital element data, call the SGP4 orbital extrapolation module to complete orbital extrapolation, and generate a nadir point time series dataset containing time, latitude, longitude, and altitude; perform data column identification on the dataset, complete time standardization and data validity verification, and remove abnormal and null sampled data.

[0017] In one implementation, two complete processing branches are split into two sets of serially executed branches. The first branch is responsible for generating the original orbital time series data, and the second branch is responsible for cleaning and regularizing the original data. The compliant and orderly sampling sequences produced by the two branches are directly supplied to the subsequent ascending node interval retrieval operation. Both processing logics are adapted to the scenario of batch satellite long-period orbit analysis. The operation rules and practical parameters of each branch are fully explained as follows.

[0018] The first branch is the acquisition of raw TLE data from SAR satellites and SGP4 orbit extrapolation calculation. The core objective of this branch is to generate continuous time-series sampling data of satellite sub-satellite points in batches by relying on publicly available TLE orbital elements and using a mature SGP4 extrapolation model. The entire calculation does not require reconstruction of the orbit extrapolation algorithm; it directly calls the existing standardized SGP4 module to complete the time-by-time point calculation.

[0019] SAR satellites with interferometric capabilities were selected as the analysis objects. Two lines of TLE orbital elements, updated daily for a 30-day period, were retrieved from the target satellite. These files comprehensively record the satellite's mean motion, orbital inclination, right ascension of the ascending node, eccentricity, and other complete orbital characteristic parameters, serving as the basic input constants for extrapolation calculations. Taking NORAD satellite 51284 as an example, the data acquisition period was defined as from November 15, 2025 to January 15, 2026, fully covering 922 orbital transit trajectories.

[0020] For the SGP4 module extrapolation configuration parameters, a unified fixed sampling interval Δt=1min is set. The module calculates the satellite spatial position step by step according to the same time step, and then projects it onto the Earth's surface to obtain the geographic coordinates of the nadir point. At each sampling time, four output indicators are calculated simultaneously: UTC sampling time, surface latitude φ, surface longitude λ, and satellite orbital altitude h. These four indicators serve as fixed core fields for the nadir point time series data, and the data storage format is uniformly time series table.

[0021] With a 30-day analysis window and a 1-minute sampling interval, a single satellite can generate more than 80,000 original sampling records. For example, the NORAD51284 satellite outputs a total of 89,280 samples, which completely restore the daily multi-circuit transit trajectory. All original sampling records are temporarily stored according to the output time sequence of the modules and directly sent to the next branch for data regularization and verification without intermediate format conversion loss.

[0022] The second branch is for the identification of nadir point time series files, time standardization and validity filtering. This branch performs three continuous processes on the original sampling table output by SGP4: field differentiation, time series unification and anomaly removal. It solves interference problems such as non-standard file headers, mixed time formats, out-of-bounds coordinate values ​​and null value distortion, and outputs a valid orbit sampling sequence with monotonically increasing time series and compliant coordinate value range.

[0023] The system has two built-in field recognition logics: keyword matching is prioritized, while fixed column order fallback is used when there is no standard English header. For keyword matching, the header strings lat, lon, time, utc, alt, and height are retrieved and bound to the latitude, longitude, UTC time, and orbital altitude columns, respectively. For fallback column order logic, if the file uses Chinese headers (such as time, latitude, longitude, and altitude), the first column is fixed as time, the second as latitude, the third as longitude, and the fourth as altitude, forcibly completing the field mapping without manual configuration. Taking the NORAD51284 satellite raw output file as an example, the file header uses Chinese identifiers; the system automatically switches to the preset column order rules, automatically binding the four types of data columns without requiring manual modification of the table header.

[0024] All time fields in the table are extracted and uniformly converted to the standard UTC timestamp format. The timestamp values ​​are then globally reordered from smallest to largest to eliminate potential local timing errors in SGP4 batch output. The original file contains various formats such as hour, minute, and second segmented text and decimal time. The standardization program uniformly converts these formats into continuous floating-point timestamps to ensure the accuracy of subsequent calculations of time differences between adjacent sampling points and to provide a continuous time reference for ascending intersection interpolation.

[0025] S102, traverses the verified nadir point time series data, identifies the zero-crossing intervals where latitude transitions from negative to positive values, and locks the ascending node interval corresponding to each orbit.

[0026] In one implementation, an ordered full-domain traversal of the effective orbit sampling sequence is completed. After data column identification, time standardization, and removal of null and outlier data, the complete sub-satellite point time-series data, arranged in ascending chronological order, is traversed throughout the entire process, adhering to the original sampling time sequence without disrupting the data's temporal arrangement structure. The latitude field value is extracted from each individual sampling record for subsequent conditional judgment. The traversal range covers all satellite transit sampling points within the analysis period, ensuring no data omissions. The NORAD 51284 interferometric SAR satellite is selected, and a 30-day analysis period is set from November 15, 2025 to January 15, 2026. The SGP4 module sampling interval is set to 1 minute. After preprocessing, a total of 89,280 effective sampling records are retained, all arranged in ascending order of UTC time. The traversal program reads the latitude values ​​of each of the 89,280 records from beginning to end, caching each latitude data separately to ensure that any two consecutive sampling points can be paired to complete the latitude zero-crossing judgment, preventing any omission of adjacent sampling combinations.

[0027] The system completes the latitude zero-crossing condition determination and single-orbit ascending node interval locking. Using two adjacent consecutive sampling records as a judgment unit, it continuously compares the latitude values ​​of the preceding and subsequent sampling points, matching the judgment criteria. When the latitude value of the preceding sampling point is less than zero, and the latitude value of the subsequent sampling point is greater than or equal to zero, the group of adjacent sampling points is designated as the interval containing the ascending node of a single orbit. Each group of adjacent sampling points that meets the judgment criteria uniquely corresponds to the equatorial transit section of one satellite orbit, thus achieving automatic segmentation of multi-orbit orbits. The divided ascending node intervals are directly transferred to the ascending node linear interpolation process. For example, NORAD satellite number 51284 uses a single group of adjacent samples, with the preceding sampling time being 2025-11-23 01:30:42, corresponding to the latitude... =-1.26°; Subsequent sampling time 2025-11-23 01:31:43, corresponding latitude =0.83°, substituting into the discriminant constraint satisfies <0 and If the value is ≥0, the preceding and following sampling points are defined as the ascending node interval of one orbit. Adjacent latitudes were compared for each of the 89,280 samples collected over 30 days, ultimately identifying 922 sets of ascending node intervals that met the constraints. Each set of intervals independently corresponds to a single equatorial transit of the satellite, and the division process did not involve orbital confusion or missed identification of ascending node intervals.

[0028] S103 performs linear interpolation calculations for each ascending node interval, and simultaneously performs longitude circumference correction during the interpolation process to obtain the time, longitude, and altitude parameters corresponding to the ascending node of each orbit.

[0029] In one implementation, the sampling points of the ascending node interval where latitude changes from negative to positive are subjected to interval locking and numerical analysis to extract the original sampling data of the first and last times, latitude, longitude, and altitude of the interval, and to construct the interpolation base dataset and the longitude discrimination logic across ±180° meridians. Each ascending node operation unit consists of two sets of original sampling data: a pre-sampling record at time i and a post-sampling record at time i+1. The program reads four fixed observation fields from the two records one by one: UTC sampling time. , Sampling latitude , Sampling longitude , Track height , After extraction, the four paired values ​​within the same interval are uniformly collected and a separate interpolation base dataset for that interval is generated. Each interval dataset is stored independently and retrieved separately during operation, so there will be no mixing of values ​​from different track intervals.

[0030] Using the analysis of NORAD satellite 51284 as a practical example, this study selected a 30-day continuous TLE extrapolation time series from November 15, 2025 to January 15, 2026. The SGP4 module sampling interval was set to 1 minute. After preprocessing, a total of 922 latitude-to-zero rising node intervals were identified. The program traversed all 922 intervals, extracting the raw values ​​of time, latitude, longitude, and altitude for each interval. An example of single-interval data extraction is provided: [Pre-sampling point...] =2025-11-23-01:30:42, =-1.26°, =178.92°, =605.92km post-sampling point =2025-11-23-01:31:43, =0.83°, =-179.65°, After the 605.91km interval was extracted, the above eight sets of values ​​were packaged to generate the interpolation base dataset specific to this orbital circle. After the batch extraction of 922 intervals was completed, 922 independent datasets were generated and cached in the memory storage area for subsequent calculation of the interpolation ratio r and calculation of the ascending node time / longitude / altitude.

[0031] Extracting each interval , After obtaining two sets of longitude values, the system performs an absolute value difference determination operation. The determination formula is: | - If the difference exceeds 180°, and the result is true, the sampling points are determined to cross the 180° east-west longitude boundary, and the interval is marked as a cross-longitude interval. In the interpolation stage, 360° longitude compensation is performed first. If the absolute value of the difference is ≤180°, it is determined to be a regular longitude interval, requiring no pre-compensation, and is directly substituted into the linear interpolation formula. The discrimination calculation is performed using the single-set interval values ​​from NORAD51284 as described above. =178.92°, The longitude difference of -179.65° is -179.65° - 178.92° = -358.57°. The absolute value of this difference, |-358.57°|, is greater than 180°. The program automatically marks this interval as a cross-meridian interval. Before interpolation, one set of longitudes is increased by 360° to achieve numerical continuity, eliminating interpolation distortion caused by longitude jumps. The entire discrimination logic is executed in a batch loop with 922 intervals. All cross-meridian intervals are marked and archived. Subsequent linear interpolation processes will prioritize reading the markings to distinguish between compensated and uncompensated interpolation processing branches. Both branches share the same linear interpolation calculation formula. ; ; ; In the formula, , This is the longitude value after 360° compensation.

[0032] The linear variation ratio of latitude within the interval, the 360° longitude pre-compensation rule, and the normalization standard for the interpolated value range are standardized to generate linear interpolation calculation rules for solving the ascending node of the adapted orbit and detailed rules for longitude orbit correction. Based on the measured orbital characteristics of latitude changing approximately linearly with time when the satellite flies over the equatorial region, the identification conditions for latitude zero-crossing intervals and the calculation logic for solving the interpolation ratio are solidified. A fixed discriminant is used for interval determination. The interpolation ratio uses a fixed calculation formula. The two sets of formulas serve as the basic calculation standards, and all latitudes passing zero are uniformly solved by applying these formulas to complete the proportional calculation.

[0033] Sampling was performed on a single ascending node interval from satellite NORAD 51284, with the previous point latitude... =-1.26°, latitude of the back point =0.83°, substituting into the proportional calculation formula, we get r=(0-(-1.26)) / (0.83-(-1.26))=1.26 / 2.09, which calculates the interpolation ratio value. This ratio is used for subsequent time, longitude, and altitude synchronization interpolation calculations.

[0034] Compare the absolute value of the difference between the longitudes of two points within the interval. When the sampling combination is determined to cross the 180° east-west boundary, before performing interpolation, a 360° overlay is first applied to one set of longitude values ​​to achieve numerical continuity, eliminating interpolation distortion caused by longitude jumps. If the absolute value of the longitude difference does not exceed 180°, the original longitude is used directly for interpolation without compensation. The longitudes before and after the interval in the same group for NORAD satellite 51284 are as follows: =178.92° =-179.65°, calculate the difference -179.65° - 178.92° = -358.57°. The absolute value of the difference is greater than 180°, so for negative longitudes, add 360°. After processing... =180.35°, interpolation calculations were performed using continuous longitude.

[0035] After all the interpolation calculations for the time, longitude, and altitude of the ascending node are completed, the longitude of the interpolated ascending node is constrained separately. If the calculation result exceeds the interval [-180°, 180°], numerical convergence is achieved by adding or subtracting 360°. Ultimately, all ascending node longitudes fall within the continuous numerical interval of -180° to 180°, ensuring that subsequent longitude closure error calculations do not encounter abrupt boundary changes. For example, if the calculated ascending longitude after interpolation for a certain interval is 186.21°, the range is closed by performing 186.21° - 360° = -173.79°; another set of interpolated longitudes is -189.43°, which is closed by performing -189.43° + 360° = 170.57°. Both corrected values ​​fall within the standard interval.

[0036] The aforementioned latitudinal linear constraints, cross-meridian compensation constraints, and interpolation range convergence constraints are all integrated and summarized, and split into two types of standardized execution files that can be used in batches. The first type of file contains the linear interpolation calculation rules, and the second type of file contains the longitude circumference correction execution details. The linear interpolation calculation rules contain five sets of fixed calculation formulas, namely, the latitude zero-crossing interval determination formula, the interpolation ratio solution formula, and the ascending node time calculation formula. Formula for calculating the longitude of the ascending node Formula for calculating the height of the ascending node The detailed rules for longitude correction include the identification conditions for crossing meridians, the sequence of 360° pre-compensation operations, and the longitude normalization processing steps after interpolation.

[0037] Using the 922 complete sets of latitude zero-crossing ascending node intervals of NORAD satellite 51284 as the batch processing objects, a complete interpolation process was executed by uniformly applying three types of standardized constraints to all intervals. The entire process consistently adopted the premise of linear variation of equatorial latitude. For each interval, the latitude discriminant was first substituted to confirm the zero-crossing attribute, and then the interpolation ratio was calculated. For each sample, the longitude difference was first compared; if the condition of crossing meridians was met, 360° numerical compensation was prioritized, and then the five sets of interpolation formulas were substituted to simultaneously solve for the ascending node time, longitude, and altitude. After all interpolation was completed, the longitude was uniformly corrected to the standard interval. The 922 intervals uniformly reused the same set of linear interpolation rules and longitude correction details. The five sets of calculation formulas remained unchanged throughout the process. During batch calculations, only the original sampled latitude, longitude, time, and altitude values ​​for each interval were replaced. The entire set of standardized documents can be directly used for batch interpolation calculations of multiple satellite orbits in the same batch.

[0038] The interpolation dataset, linear interpolation calculation rules, and longitude correction execution details are integrated. Based on the orbital orbit ascending node solution logic, cross-boundary longitude pre-compensation correction, linear interpolation calculation, and result range normalization are completed to generate standardized ascending node time, longitude, and altitude parameters for each orbital orbit. The linear interpolation calculation rules include... ; ; ; ; .in, Let be the latitude of the i-th sampling point; This is the latitude linear interpolation weighting coefficient, with a value ranging from 0 to 1; The standard UTC time is the time at the ascending node of the equator. The observation time for the previous sampling point; The observation time for the subsequent sampling points; This is the interpolation scaling factor; The longitude of the ascending node obtained by interpolation; The longitude is sampled before and after the interval; The altitude of the ascending node; , This refers to the satellite orbital altitude before and after the interval. The first step is to perform a 360° pre-compensation operation for the longitude of the sampling points across ±180° meridians. This involves first reading the longitudes of two points before and after the same interval in the interpolation base dataset. , Calculate the absolute value of the difference between two sets of longitudes, when | - | At 180°, any set of longitudes within the interval is selected and superimposed with 360° to complete the continuous numerical conversion, eliminating the numerical jump caused by the intersection of east and west meridians; if the absolute value of the difference is not greater than 180°, the original longitude value is directly used without compensation. For example, sampling a certain ascending node interval of the NORAD satellite with serial number 51284, the pre-sampled longitude... =178.92°, post-sampling longitude =-179.65°, the difference between the two is -358.57°, the absolute value of the difference is greater than 180°, and 360° is superimposed on the latter longitude, after processing =180.35°, and the two corrected sets of longitudes are used in subsequent interpolation calculations.

[0039] The second step performs a complete linear interpolation operation. It retrieves all values ​​for the standardized latitude, corrected longitude, original time, and original altitude of the interval, and substitutes them into five sets of fixed interpolation formulas to simultaneously solve for the ascending node time, longitude, and altitude. The interpolation ratio *r* is uniquely determined by the latitude samples before and after the interval. The ratio is calculated only once for the same interval and reused throughout the entire calculation process for time, longitude, and altitude. Numerical example: the latitude sampled before the interval. =-1.26°, post-sampling latitude =0.83°, substituting into the proportional calculation formula r=(0-(-1.26)) / (0.83-(-1.26))=1.26 / 2.09; Precedence time =2025-11-23-01:30:42, Post-time =2025-11-23-01:31:43, substitute r into the time formula to calculate the UTC time corresponding to the ascending node; simultaneously substitute the corrected longitude value and the original orbital altitude to solve for the longitude and orbital altitude of the ascending node of this circle in one go.

[0040] The third step involves normalizing the interpolated longitude range. Since the ascending node longitude obtained in the second step may exceed the range of [-180°, 180°], interval convergence is performed on the interpolated longitude. >180°, execute = -360°; if <-180°, execute = +360°, ultimately constraining the longitude of the ascending node to the range of -180° to 180°. Numerical example: One interpolation directly outputs the longitude of the ascending node as 186.21°, and convergence is achieved by performing 186.21°-360°=-173.79°; another interpolation outputs a longitude of -189.43°, and convergence is achieved by performing -189.43°+360°=170.57°. Both corrected longitudes fall within the standard range.

[0041] Using the 922 complete sets of ascending node intervals with latitude zero crossings from the NORAD satellite 51284 as the batch calculation objects, each interval independently retrieves its own dedicated interpolation base dataset and processes them in a fixed three-step cyclical sequence: compensation, interpolation, and normalization. Each interval first completes the meridian crossing determination and 360° pre-compensation, then substitutes the entire set of linear interpolation formulas to simultaneously solve for the three ascending node parameters. Finally, the interpolated longitude values ​​are converged, and a standardized set of ascending node time, longitude, and altitude parameters is generated separately for each interval. After all 922 intervals have completed the three-segment continuous calculations, all ascending node parameters are summarized in the order of orbit generation time to form a temporally ordered full-orbit ascending node dataset. This complete dataset is directly used as input data to the longitude closure error calculation and processing stage for subsequent error solving and two-dimensional plane baseline conversion.

[0042] S104 uses the longitude of the ascending node of the first circle as the reference value, calculates the longitude closure error of the remaining ascending nodes for each circle, and normalizes the error range to the continuous interval of [-180°, 180°].

[0043] In one implementation, benchmark assignment and closure error modeling are performed on the longitude dataset of all ascending nodes in the orbital cycles. This is achieved through a three-step progressive architecture: benchmark longitude selection, solving for the original error on a cycle-by-cycle basis, and normalizing the out-of-limit value range. The longitude corresponding to the first ascending node record in the dataset is selected according to the chronological order of orbit generation. As a globally unified benchmark reference value, the complete ascending node longitude time series is split into two independent data sets. The first set stores only a single benchmark longitude. The first group is fixed throughout and is used only as a reference for difference calculations; the second group stores the longitudes of the ascending nodes for all orbital cycles except the first. As a variable used separately for comparison in each orbit, the two sets of data cannot be mixed or cross-referenced; each must have its own independent access boundaries. For example, the NORAD satellite 51284 generated 922 sets of standardized ascending node longitudes during a 30-day analysis period, with the first time series showing the ascending node longitude of the first orbit. =23.61°, a separate benchmark dataset was generated; the remaining 921 orbital longitudes were all included in the dataset to be calculated. During the batch calculation, each set of longitudes to be calculated was fixedly paired with 23.61° for difference calculation, and there was no replacement of benchmark values ​​or mixing of multiple benchmarks.

[0044] Retrieve baseline dataset with fixed longitude Longitude of the ascending node to be calculated for a single circle The original closure error is calculated using the fixed difference formula, which is as follows: The algorithm iterates through all orbit records in the dataset to be calculated, substituting each record into the formula to obtain the original longitude offset values ​​without range adjustment. The original error retains its complete positive and negative characteristics and is used to characterize the east-west offset direction of the ascending node relative to the reference. For example, the longitude of the ascending node of the 119th orbit of NORAD satellite 51284. =-131.39°, substitute into the formula =-131.39°-23.61°=-155.00°; Longitude of the ascending node of the other orbit. =204.82°, the original closure error is 204.82°-23.61°=181.21°, and the two sets of original errors directly enter the value range correction step.

[0045] The effective longitude range is set to [-180°, 180°]. An interval determination is performed on all the original closure errors obtained in the second step, with two fixed compensation rules: If... ,implement ; ,implement After 360° compensation, all error values ​​are converged to the continuous interval [-180°, 180°]. The converged values ​​are normalized longitude closure errors, which are then uniformly summarized and flowed into the two-dimensional sub-satellite point equivalent plane baseline conversion stage. For example, if the original error calculated in the previous step is 181.21°, which meets the condition of being greater than 180°, then 181.21°-360°=-178.79° is calculated. If the original error of a certain orbit is -231.46°, which meets the condition of being less than -180°, then -231.46°+360°=128.54° is calculated. Both sets of corrected values ​​fall within the standard value range, and there is no longitude boundary jump problem.

[0046] Thirty days of orbital data from NORAD satellite 51284, from November 15, 2025 to January 15, 2026, were selected. The interpolation process output 922 sets of time-ordered, standardized ascending node longitudes, all of which were aggregated into a complete input dataset. Batch processing followed a three-stage, fixed-order sequential execution: The first stage extracted the first orbit's 23.61° as a baseline, splitting the dataset into baseline and datasets to be calculated; the second stage iterated through the 921 orbits, substituting each into the difference formula to solve for the original closure error; the third stage performed range determination and 360° compensation correction on all original errors, generating 922 sets of normalized longitude closure error time-series data. This complete dataset was directly transferred downstream for angle-to-meter baseline conversion calculations. The entire batch processing workflow maintained the same order of the three operations, eliminating issues of missed orbits or mixed baselines.

[0047] The baseline anchoring was completed according to the order of the orbital orbits, and the longitude of the ascending node of the first orbit was determined. A global baseline reference value is established, and the datasets are divided into a baseline longitude dataset and a dataset of ascending node longitudes for each orbit to be calculated. The calculation boundaries and input purposes of the two types of data are clearly defined. The selection of the orbital ascending node baseline is entirely based on the actual sequence of orbital passage, without using random orbits, intermediate orbits, or last orbits as reference baselines. The ascending node longitude corresponding to the first orbit in the time sequence is fixed and selected. As a globally unified reference value, this benchmark value remains constant throughout the entire process of calculating orbital errors and does not change with the switching of calculation cycles. All subsequent orbital ascending nodes can only be paired with this fixed value for difference calculation, and multiple sets of comparison benchmarks are not set.

[0048] All collected time-series ascending node records are split into two isolated datasets. The two datasets are stored in independent partitions, and their computation and retrieval channels are separate, eliminating cross-data retrieval and mixed computation. For the baseline longitude dataset: only the longitude of the first orbit's ascending node is stored. The data is used solely as a fixed reference constant for difference calculation, and is read only once during each lap error calculation, without any cyclic updates or replacements. For the dataset of laps to be calculated: it stores the longitudes of the ascending nodes of all orbits except the first lap in the time series. Each record corresponds independently to a satellite transit node. During the calculation, each record is retrieved individually in chronological order and paired with a fixed reference to complete the difference calculation.

[0049] After dividing the datasets into two categories, their respective computational boundaries are clearly defined: the baseline dataset does not participate in the single-lap iteration loop, but is only used as a global constant; the dataset to be calculated fully participates in the lap-by-lap error calculation loop. The two types of data are not allowed to be interchanged, thus avoiding computational errors such as mixing baselines and arbitrarily changing reference values ​​at the data reading level. After pairing the baseline and the dataset to be calculated, the original longitude closure error is solved using a fixed formula: In the formula Represents the longitude of the first benchmark ascending node. This represents the longitude of the ascending node of the orbit to be calculated on the kth orbit. This represents the original closure error without range compensation.

[0050] A 30-day analysis period was selected for NORAD satellite 51284, from November 15, 2025 to January 15, 2026. After ascending node interpolation, a total of 922 time-series ordered ascending node longitude records were generated, all arranged from earliest to latest UTC time. The longitude of the first ascending node of the first orbit was extracted. =23.61°, stored separately in the baseline dataset, which retains only this single value throughout the entire dataset, and remains unchanged throughout the batch calculation; after removing the first orbit record, the remaining 921 longitudes of the ascending nodes of each orbit are all included in the dataset to be calculated, and are completely preserved in their original time order.

[0051] Iteratively reads the values ​​of each ascending node in the dataset to be calculated, and reads each set of values ​​in turn. The original error was calculated by substituting the values ​​of the fixed reference point 23.61° into the difference formula. A specific calculation example for a single lap is as follows: Longitude of the ascending node of the orbit in lap 119. =-131.39°, substitute into the formula to calculate. =-131.39°-23.61°=-155.00°; the ascending node of the other orbit. =204.82°, the original error was calculated. =204.82°-23.61°=181.21°.

[0052] For the two sets of longitude datasets, a standardized closure error calculation logic was built, first according to the formula. The original longitude closure error is solved circle by circle, and then ±360° compensation correction is performed on errors exceeding the [-180°, 180°] interval. If ;like This step completes the output of the normalized longitude closure error. The pre-input for this step consists of two independent datasets: one is a baseline dataset storing only the fixed ascending reference longitude of the first orbit, and the other is a dataset containing the longitudes of all remaining ascending nodes arranged sequentially according to the orbit's passage time. These two types of data were already separated in the previous step, with the baseline longitude... The entire route remains fixed, longitude to be calculated. Data was retrieved individually, one track at a time, and the two sets of data were paired one-to-one for difference calculation. The original longitude closure error was solved using a fixed difference formula: In the formula, Represents the longitude of the ascending node in the first leg. This represents the longitude of the ascending node corresponding to the orbit to be calculated in the kth orbit. This is the original closure error without range correction. This value retains the positive and negative signs and directly reflects the east-west offset direction of the current orbit relative to the reference ascending node.

[0053] Taking the measured data from NORAD satellite 51284 as an example, the reference longitude of the first orbit within the entire ascending node time series of this satellite is... =23.61°, where the longitude of the ascending node of the 119th orbit is... =-131.39°, substitute into the formula to calculate the original closure error: =-131.39°-23.61°=-155.00°. The absolute value of this original error is less than 180°, so no compensation is needed and it can be used directly as the normalized error. There is also the longitude of the ascending node of the orbit. =204.82°, substituting into the formula, we get the original error: =204.82°-23.61°=181.21°, which exceeds the range of [-180°, 180°] and needs to be processed for compensation.

[0054] The original closure error obtained for each set of solutions Numerical range determination is carried out, and two fixed correction execution rules are defined: if >180°, perform numerical correction: = -360°; if <-180°, perform numerical correction: Only when the original error falls within the range of -180° to 180°, no additional or subtractive compensation operations are required; the original error is directly equivalent to the normalized closure error. All correction operations are calculated only for a single error value and do not change the original records of the reference longitude or the original ascending node longitude of each orbit.

[0055] The following are two complete calculation examples of out-of-limit values. Example 1: The original closure error of a certain orbital loop is 205°, which meets the criterion of being greater than 180°. According to the first rule, subtraction compensation is performed: 205° - 360° = -155°. After compensation, the value -155° is within the standard value range of [-180°, 180°], which is an effective normalized longitude closure error. Example 2: The original closure error of a certain orbital loop is -231°, which meets the criterion of being less than -180°. According to the second rule, addition compensation is performed: -231° + 360° = 129°. After compensation, the value 129° falls within the standard value range and can be directly transferred to the next process.

[0056] After completing the single-circle error determination and compensation operations, the resulting values ​​are uniformly named Normalized Longitude Closure Error. The normalized errors corresponding to all orbital circles within the analysis period are then collected sequentially to generate a complete time-series error dataset. This dataset serves as the sole standard input, fully transferred to the two-dimensional nadir point equivalent plane baseline conversion stage for angle-to-meter length conversion calculations. When the entire process batch-processes 922 groups of NORAD satellite orbital circles (number 51284), all original errors undergo interval verification. All normalized errors, after compensation or direct retention, do not exceed the boundary values ​​of [-180°, 180°], thus avoiding baseline calculation distortion issues caused by crossing ±180° meridians.

[0057] The NORAD satellite with serial number 51284 has a total of 922 sets of standardized ascending node longitude records. The first orbit's longitude of 23.61° is used as the global baseline. The remaining 921 sets of ascending node longitude records are paired sequentially to solve for the original errors. Each set of original errors is subject to interval judgment. Values ​​exceeding the limit are uniformly compensated with 360° addition and subtraction. After all processing is completed, the 922 sets of ordered normalized errors are summarized. The complete dataset has no missing orbits and no mixed references. It can be directly input into the baseline dimension conversion calculation process.

[0058] S105, combined with the Earth's equatorial radius parameter, converts the normalized ascending node longitude closure error into a meter-level two-dimensional sub-satellite equivalent plane baseline.

[0059] In one implementation, baseline dimension transformation modeling is performed on the normalized longitude closure error dataset, employing a three-step progressive architecture of radian conversion, radius coupling, and plane baseline solution. The sole input to this model comes from the output of the longitude closure error processing step. All normalized longitude closure errors need to be uniformly collected to generate a complete time-series error dataset. Each error value in the dataset undergoes ±360° compensation correction in the preceding steps. All values ​​converge within the range of [-180°, 180°], with no out-of-bounds angles greater than 180° or less than -180°. The dataset's arrangement order is consistent with the orbital transit UTC time, ensuring the time series remains intact and there are no missing error records corresponding to the orbit. For example, after ascending node interpolation and closure error normalization, the NORAD satellite with serial number 51284 generated a total of 922 time-series ordered normalized error records. All error values ​​were closed in their ranges. The normalized error for a single loop was -155° and the normalized error for another loop was 129°. Both types of values ​​were within the standard range and there were no boundary jumps.

[0060] A dedicated dimensional conversion calculation model for angle-to-meter conversion is built based on a time-series error dataset with compliant value ranges. The model is fixedly divided into three independent calculation processes, which are subject to irreversible sequential execution constraints. All data calculations in the previous process must be completed before proceeding to the next process; changing the order of calculations and parallel calculations across multiple processes are prohibited. Data output from each process is only transmitted unidirectionally to the next process as dedicated input, with no data backflow. The entire model only performs conversion calculations for orbital equatorial plane closure deviations, without incorporating vertical components of orbital altitude or calculation logic related to three-dimensional spatial distances; it only outputs a two-dimensional sub-satellite equivalent plane baseline. The functions of the three processes are as follows: First process: Angle-to-radian conversion, converting the degree-normalized longitude closure error into radian units. Second process: Radius coupling calculation, substituting the fixed radius of the Earth's equator to complete the angle-to-length conversion. Third process: Planar baseline numerical output, generating a two-dimensional equivalent baseline in meters with positive and negative signs.

[0061] Specifically, the first section is the conversion of angles and radians, using the normalized closure error of a single loop. For input, complete the angle unit conversion. The conversion formula is taken from the recorded standard formula: In the formula, Represents the angular offset in radians. This is the normalized longitude closure error in degrees. For example, the normalized error of a certain orbit of NORAD satellite number 51284. =-155°, substitute into the formula to calculate the radian value: The converted radian value serves as the sole input for the second segment of the calculation.

[0062] The second part involves the coupling calculation of the Earth's equatorial radius. Using a fixed equatorial radius constant R = 6378137m, the radian offset obtained in the first part is multiplied by the radius to complete the conversion from angle to plane length, with a complete conversion formula provided: In the formula, The baseline is the equivalent plane of the untrunculated two-dimensional sub-satellite point, with the unit uniformly in meters. The constant R remains unchanged throughout the entire orbit, and this value is reused for all orbits without being changed with orbital switching. For example, the radian value corresponding to the -155° normalized error is coupled with the equatorial radius of 6,378,137 meters to calculate the final baseline value with a negative sign, where the negative sign represents the offset azimuth relative to the reference orbit.

[0063] The third segment is the output of the two-dimensional sub-satellite point equivalent plane baseline, obtained through a two-step coupled calculation of radians and radius. This data is directly used as the final baseline parameter for a single orbit, retaining the complete positive and negative signs in the output. Positive and negative numbers represent the east-west plane offset of the ascending node relative to the reference orbit, respectively, without absolute value processing. The entire output baseline only characterizes the degree of equatorial plane closure and does not reflect the vertical deviation of the satellite's high and low orbits. Numerical calculation example: A -155° error, calculated using the entire formula, yields a baseline value of -1826m. This value is the equivalent plane baseline of the two-dimensional nadir point for this orbital orbit and is directly stored in the baseline time-series dataset for the local trough identification process.

[0064] Using 30-day orbital analysis data from NORAD satellite 51284 as the batch processing object, all 922 sets of compliant and normalized longitude closure errors were collected. Each set of error data was processed cyclically in a fixed sequence of three stages: radian conversion, radius coupling, and baseline output. Batch execution details: Time-series normalized errors were retrieved for each orbit. The first step was to uniformly convert degrees to radians; the second step was to substitute a fixed equatorial radius of 6,378,137 meters and perform coupled multiplication; the third step was to directly retain the signed meter-level baseline values, completing the generation of single-orbit baseline records. After all 922 sets of orbital cyclic calculations were completed, all baseline values ​​were summarized in chronological order to generate a complete two-dimensional equivalent baseline time-series dataset. This complete dataset was directly passed down as the input data source for local trough identification and 2000m threshold filtering. Batch Operation Instructions: The 922 sets of orbits reuse the same set of conversion formulas and the same equatorial radius constant throughout the entire process. The three-segment processing order is fixed throughout the entire process. There is no batch processing method that skips a segment or changes the calculation steps. All errors have undergone complete dimensional conversion, and there are no orbit records with skipped calculations.

[0065] Based on the mathematical rules of angle-to-length conversion, parameter matching is implemented, and the normalized longitude closure error and the Earth's equatorial radius are divided into angle input subsets and length reference subsets, clarifying the calculation boundaries and unit matching requirements for the two types of parameters. The first type is the angle input subset, which stores the full-time-series longitude closure error after previous normalization processing. All numerical units are uniformly set to degrees, and all values ​​have undergone ±360° range compensation, converging to the [-180°, 180°] interval. This subset can only be retrieved and used in the first stage of the radian conversion process. Radius coupling and baseline output segments do not read the original angle data of this subset, but only receive the converted radian results as input.

[0066] The second category is the length reference subset, which retains only one fixed constant: the Earth's equatorial radius, with a standard value of 6,378,137 meters. The unit of measurement is fixed in meters, and its value does not change throughout the entire process regardless of satellite type, analysis period, or orbital revolution. This subset is only allowed to be retrieved in the second stage of the dimension conversion radius coupling process. The radian conversion segment does not call this length constant; it is only used for paired multiplication operations after the radian values ​​have been converted.

[0067] The two sets of parameter subsets are set to completely isolate the data reading boundaries. The radian conversion process only calls the angle subset and does not read the equatorial radius. The radius coupling process simultaneously calls the radian conversion result and the length constant. The baseline output segment no longer reads the two types of original subset data. The two types of original parameters do not directly participate in the same round of multiplication operation. They are only paired and coupled after the radian unit conversion is completed. All length-type output results are forced to use meters as the unit of measurement throughout the process. There is no situation where mixed units such as kilometers and degrees are involved in the baseline calculation, which avoids the distortion of baseline values ​​caused by unit conversion omissions.

[0068] Degree to radian conversion formula In the formula, Representative angle input subset, single-circle normalized longitude closure error, in degrees; This represents the radian offset after conversion. Formula for calculating the baseline coupling of the equivalent plane at the 2D sub-satellite point. In the formula, R represents the length reference subset, which incorporates the Earth's equatorial radius constant of 6378137 m. The final signed plane baseline, in meters.

[0069] Complete parameter matching and conversion calculations were performed using single-orbit data from NORAD satellite number 51284: the normalized longitude closure error of this orbit. =-155°, included in the angle input subset; the length reference subset is fixed and read as a constant R=6378137m.

[0070] The first radian conversion is as follows: Substitute -155° into the radian conversion formula. The angular offset in radians is obtained. This intermediate result is only passed down to the second coupling process; it is not stored separately and does not participate in other segment calculations. The second radius coupling calculation is as follows: the radian result and the equatorial radius of 6,378,137 meters are substituted into the baseline complete formula for calculation. The calculation is completed and outputs the two-dimensional equivalent baseline value in meters directly. The entire calculation does not use other length units such as degrees or kilometers. The two types of parameter derived values ​​are paired and multiplied only in the coupling step.

[0071] When performing batch conversion on all 922 sets of orbit normalization error datasets for NORAD satellite number 51284, the angle input subset is retrieved one by one according to the orbit time sequence. Numerical values ​​and length reference subsets are fixed and reused as a constant of 6,378,137 meters throughout the entire process. Each group of tracks follows a fixed operation logic of "first completing the radian conversion separately, and then performing radius pairing and coupling". There is no erroneous operation method of directly multiplying the original angle with the radius. The output unit of all track baselines is unified in meters. The batch-generated baseline time series datasets are completely fed into the local trough identification and screening process.

[0072] The equivalent baseline conversion calculation logic is built for the two types of parameters, according to the formula. The process completes angle-to-radian conversion and equatorial radius coupling calculations, outputting a meter-level two-dimensional nadir equivalent plane baseline. Paired standardization calculations are performed using the single-cycle normalization error of the angle input subset and the equatorial radius constant of the length reference subset. The first step involves a degree-to-radian conversion to normalize the angle units. The second step multiplies the converted radian value by the Earth's equatorial radius value to complete the length coupling calculation. These two steps are executed sequentially and cannot be separated or interrupted. After the calculations are complete, a meter-level two-dimensional nadir equivalent plane baseline with positive and negative signs is generated. This value only represents the relative orbital offset within the satellite's equatorial plane and is not included in calculations related to vertical deviation in satellite orbital altitude or three-dimensional spatial distance. All orbital cycle baseline values ​​are compiled to form a complete baseline time series. This complete sequence is directly transmitted to the next step, local trough identification processing, as the original input data for baseline extremum determination.

[0073] For example, the NORAD satellite number 51284 has a single orbit normalization error of -150°. The first step is to convert the corresponding radian value, and the second step is to couple the calculation with the equatorial radius of 6,378,137 meters. The final output is a signed meter-level baseline value of -1,826 meters. This value is only used to determine the degree of orbital plane closure and does not participate in the calculation of vertical deviation. After the batch output of 922 complete baseline sequences, it is directly used for baseline local minimum value retrieval and screening.

[0074] S106. Local troughs are identified in the baseline numerical sequence of all orbital cycles to screen out candidate re-orbit cycles corresponding to local minimum values ​​of the baseline. A fixed threshold of 2000m is set to filter the candidate re-orbit cycles, and samples with baseline values ​​not exceeding the threshold are retained to obtain an effective short baseline re-orbit window.

[0075] In one implementation, combining the operational requirements of SAR short baseline screening with the baseline sequence extreme value discrimination logic, a mechanism for local trough determination, fixed threshold filtering, and candidate window extraction is introduced to determine a standardized processing scheme for baseline sequence traversal, minimum value identification, threshold screening, and duplicate orbit window output. Using a complete time-series baseline dataset as the processing object, all 922 sets of baseline records are traversed from beginning to end according to the order of orbit generation. No orbital cycles are omitted during the traversal process, and the absolute value of each single-cycle baseline is retrieved and stored for subsequent size comparison. For the first and last orbitals of the sequence, where there are no adjacent baselines on one side, the ternary comparison condition is not met, and these are directly marked as not participating in the minimum value determination; only the middle orbitals can simultaneously retrieve the baseline values ​​of the previous, current, and next cycles for the trough discrimination stage. For example, the complete baseline sequence of the NORAD satellite number 51284 contains 922 records. Orbit 1 is the first record in the sequence and Orbit 922 is the last record. When traversing and reading, only the two baseline values ​​are stored, and there is no comparison data on the other side, so the minimum determination is skipped directly. The baselines of the front and rear orbits such as Orbits 119, 238, and 357 can be retrieved simultaneously to enter the next adjacent value comparison process.

[0076] A single mid-section track undergoes a ternary absolute value comparison operation, using a fixed discrimination formula: In the formula, Represents the previous two-dimensional equivalent baseline. Represents the current baseline. This represents the baseline of the next orbit. If both sets of less than or equal to conditions are met, the current orbit is marked as a local trough candidate orbit and stored in a temporary candidate set; if the absolute value of either baseline is less than the current baseline, it is directly removed from the candidate list and does not proceed to the threshold verification process. For example, the baseline of the 119th orbit of NORAD satellite number 51284. = -357.05m, absolute value 357.05m; the absolute value of the baseline of the first circle is 412m, and the absolute value of the baseline of the second circle is 455m, which satisfies the discrimination formula constraint and is marked as a candidate for trough; the absolute value of the baseline of the 238th circle is 455.89m, and the absolute value of the baseline of the second circle is 321m, which does not satisfy the double minimum condition and is directly removed from the candidate set.

[0077] Based on the general constraints of SAR interferometric imaging, a distance screening standard is constructed, and a fixed threshold is selected. =2000m, with an adjustable threshold range of 1000m to 3000m to adapt to the observation needs of satellites in different wavebands. The absolute value of the baseline of all candidate wave troughs is compared with the threshold, dividing the orbit samples into two categories: specifically, baselines with an absolute value greater than 2000m are judged as false wave invalid orbits and directly discarded; baselines with an absolute value less than or equal to 2000m are judged as valid orbits meeting the short baseline requirement for interferometric imaging and retained for aggregation. For example, the absolute value of the baseline of the 119th orbit is 357.05m, which is less than the 2000m threshold and is retained as a valid repeated orbit sample; the baseline of a certain mid-segment candidate orbit is 2130m, exceeding the threshold standard, and is judged as a false wave trough and directly eliminated.

[0078] After trough identification and threshold screening, all remaining track records are uniformly summarized in chronological order to form a standardized short baseline heavy track window dataset. The dataset fully carries the corresponding track cycle number, baseline value, transit time, and track altitude parameters. The complete dataset is directly transmitted to the downstream results generation stage as the original input material for reports and trend analysis charts.

[0079] The 30-day orbit analysis results of NORAD satellite 51284 were selected. The baseline conversion process output 922 sets of time-ordered two-dimensional equivalent baselines as input. Batch cyclical calculations were performed in a fixed sequence of four steps: traversal retrieval, trough identification, threshold screening, and window aggregation. Each of the 922 baseline values ​​was read, distinguishing between the first and last orbits and the middle orbits, and selecting the middle orbits that met the ternary comparison criteria. All middle orbits were compared using the minimum value formula to generate a temporary trough candidate set. Each candidate baseline was compared with the 2000m optimal threshold, and false troughs exceeding the threshold were eliminated. All compliant orbits were summarized to generate a standardized duplicate orbit window dataset. The entire batch calculation process maintained the same order of the four processing steps. All 922 baselines underwent two-level judgment and screening, with no missed orbits and no parallel processing across steps. The final aggregated valid duplicate orbit data was used for two subsequent tasks: exporting structured data tables and plotting baseline change curves.

[0080] A global traversal and local trough discrimination are performed on the two-dimensional equivalent baseline sequence of the entire orbital cycle to extract the data that meet the requirements. The absolute values ​​of the baselines of adjacent preceding and following orbits are both greater than the local minimum candidate orbit number of the current orbit. For any orbit of the k-th orbit in the middle of the sequence, the absolute value of the baseline of the previous orbit is retrieved synchronously. The absolute value of the current baseline Absolute value of the next circumference baseline The three sets of values ​​are compared using a fixed formula: Only when both sets of less than or equal to conditions are met simultaneously is the current orbital circle determined to be a local trough candidate and included in the set of baselines to be screened; if the absolute value of any adjacent baseline is less than the absolute value of the current circle's baseline, it is determined that it does not meet the local minimum feature and is directly removed, no longer participating in the 2000m threshold screening step.

[0081] For example, NORAD satellite 51284 generated 922 ordered baseline records during a 30-day analysis period. Orbit 119 belongs to the middle section of the sequence, and the absolute values ​​of the corresponding three baseline sets were retrieved: [Previous Orbit] =412m, Current circle =357m, back loop =455m. Substituting the value into the judgment formula, 357≤412 and 357≤455, both constraints are satisfied. The 119th circle is marked as a candidate orbit for a local trough and stored in the candidate baseline set.

[0082] The same satellite's 238th orbit was also a mid-course orbit. The absolute value of the preceding baseline. =455m, rear circle baseline =321m, the current baseline of 455 meters is greater than the rear circle's 321 meters, which does not satisfy the condition. ≤ If a candidate is determined to lack local minimum features, it is removed directly from the candidate set and does not proceed to the next threshold screening process.

[0083] The first track in the baseline sequence (circle 1) only has the baseline of the next circle, and there is no baseline value of the preceding track; the last track in the sequence (circle 922) only has the baseline of the previous circle, and there is no baseline value of the following track. Neither type of track can complete the three sets of baselines (front, middle and back) to complete the ternary comparison operation, and does not meet the data prerequisites for local trough determination. After identifying the first and last tracks during the traversal process, the comparison step is skipped directly and the candidate baseline set is not entered.

[0084] The complete baseline sequence of the NORAD satellite 51284 contains a total of 922 orbit records. In the first orbit, only the baseline of the next orbit can be read, and there is a lack of preceding comparison data, so the trough determination process is terminated directly. The 922nd orbit is the last orbit of the entire time series, and only the baseline of the previous orbit exists. There is no baseline value corresponding to the subsequent sampling, so the ternary size comparison is not carried out. Neither type of orbit will enter the subsequent 2000m threshold screening process.

[0085] After the traversal and judgment process is completed, all intermediate track cycles that pass the double minimum constraint are uniformly summarized and stored to generate an independent set of candidate baselines to be screened. Each record in the set carries the corresponding track cycle number and complete baseline value. The entire set serves as the exclusive input data for the 2000m baseline threshold filtering process. The tracks within the set maintain their original chronological order and are not reordered.

[0086] The complete batch execution process is as follows: A complete traversal judgment is performed on the 922 sets of baseline records for NORAD satellite number 51284: all baseline values ​​are read from beginning to end, distinguishing between the first and last orbits and the middle orbits; the 920 middle orbits are substituted into the ternary judgment formula one by one to complete the numerical comparison; the boundary orbits of orbit 1 and orbit 922 are removed, along with orbits such as orbit 238 that do not meet the minimum condition; all orbits that meet the double less than or equal to constraint are retained, and a candidate baseline set to be screened is generated in chronological order and passed to the next threshold screening process.

[0087] A 2000m baseline screening threshold system was constructed based on SAR interferometry optimization constraints to distinguish between two types of sample boundaries: invalid loops exceeding the threshold and effective short base loops within the threshold, satisfying the following conditions: The condition is that the compliant double orbit window extraction is completed. After completing the identification of local troughs, all candidate orbit baseline records are summarized. For each group of candidate baselines, absolute value conversion is performed. The converted values ​​are compared with the preferred threshold to classify the orbit samples into two categories. The classification rules are as follows: if... >2000m: This trough is a pseudo-trough; the corresponding orbital circle does not meet the baseline conditions for interferometric imaging and is therefore directly discarded and will not be included in the candidate dataset for repeated orbits. If ≤2000m: This orbital circle is a valid short baseline sample, and all associated parameters of the orbit are completely preserved.

[0088] The compliant orbits selected through two rounds of screening are uniformly collected and organized to generate a standardized candidate orbit window dataset. Each record in the dataset is synchronously bound to the orbit circle number, satellite orbit collision time, orbit ascending node altitude, and the original value of the two-dimensional equivalent baseline with signs. The dataset is transmitted only unidirectionally to the next output generation process, without any data link for reverse transmission to the baseline identification and baseline conversion stages. The dataset is used for two processing flows: generating structured indicator tables and plotting baseline change trends.

[0089] Threshold comparison was conducted using 922 sets of baseline sequences from NORAD satellite number 51284. The baseline corresponding to orbit 119 was selected. = -357.05m, take the absolute value =357.05m, which is less than the preferred threshold of 2000m, and is therefore determined to be a valid short baseline orbit. The complete data includes orbit number 119, re-orbit time 2025-11-23 01:31:27, orbit altitude 605.918km, and baseline -357.05m to the candidate dataset. A certain mid-section orbit was marked as a candidate after trough identification. The baseline value is 2130m. The absolute value of 2130m is greater than 2000m, so it is determined to be a false trough. All records of this orbit are directly discarded and will not be included in the subsequent results processing.

[0090] For long-band SAR satellite observation services, the imaging allows for a wider baseline range. The screening threshold can be reset within an adjustable range of 1000m to 3000m. Taking the 2800m adjustment scenario as an example: the global comparison standard is changed from 2000m to 2800m, and a new round of comparison and screening is performed on the absolute values ​​of all local valley candidate baselines. Orbits with a value greater than 2800m are identified as false valleys, while orbits with a value less than or equal to 2800m are retained as short baseline windows. The adjusted dataset is adapted to the planning requirements of long-band interferometric observation missions. Similarly, for short-band observation scenarios, the threshold can be lowered to values ​​within the range of 1000m, 1500m, etc., to adapt to narrow baseline imaging constraints.

[0091] S107 integrates the equivalent plane baseline calculation results with the filtered short baseline re-rail window information, and uniformly outputs the re-rail time, re-rail cycle, baseline value, and track height data, while simultaneously generating a track cycle-baseline change analysis curve.

[0092] In one implementation, an integrated approach is adopted, encompassing baseline statistical collection, duplicate track information fusion, visualization, and simultaneous output of multiple results. This approach unifies the output process, matches baseline data collection granularity to the time sequence of the orbital loops, sets the execution order of structured numerical output and baseline trend chart drawing according to business priorities, and categorizes output into two types: standardized data tables and baseline analysis charts. All orbital records are traversed sequentially by orbital time, uniformly incorporating four types of original information: orbital loop number, two-dimensional equivalent baseline, ascending node altitude, and transit UTC time, thus constructing a unified time-series index. For example, during the analysis period of NORAD satellite 51284, a total of 922 orbital baselines were collected. After threshold screening, three sets of compliant multiple orbit records (numbers 119, 238, and 357) were selected. During the aggregation phase, the original parameters of the three windows were fully recorded: orbit 119: transit 2025-11-23 01:31:27, altitude 605.918km, baseline -357.05m; orbit 238: transit 2025-12-01 01:31:27, altitude 605.898km, baseline -455.89m; orbit 357: transit 2025-12-09 01:31:24, altitude 605.761km, baseline 878.43m. The 922 sets of complete orbital information were archived simultaneously, providing a time-series reference for subsequent parameter matching and mapping.

[0093] Based on the aggregated time-series index, the overlap interval period is calculated for each group of qualified overlap rail windows, and the five parameters of circle number, time, period, baseline, and track altitude are bound to a complete single record. Calculation example: Circle 119 is the first valid overlap rail, corresponding to a period of 8 days, 0 hours, and 0 minutes; the time difference between Circle 238 and Circle 119 is calculated to have a period of 16 days, 0 hours, and 0 minutes; the interval between Circle 357 and Circle 238 is 23 days, 23 hours, and 59 minutes. All parameters are uniquely bound to their corresponding tracks, and there are no mismatch issues.

[0094] The horizontal axis of the plot is uniformly labeled with orbital revolution numbers (arranged in ascending order from 1 to 922), and the vertical axis represents the baseline in meters. The plot includes three types of fixed plotting elements: a continuous fluctuation curve of the entire orbital baseline, a 2000m horizontal threshold reference line, and a special marker symbol for the overlapping orbit window. The three sets of overlapping orbit points (119, 238, and 357) are labeled at their corresponding positions on the horizontal axis. The plot format is consistent across all satellites, supporting horizontal comparison of multiple plots. A structured data table of overlapping orbit indicators is generated first, with fixed fields including overlapping orbit number, transit time, overlapping orbit period, baseline, and orbital altitude. After the table is generated, a baseline analysis curve is simultaneously output. Both types of files are archived as a set and not saved separately.

[0095] Taking the NORAD51284 satellite as an example, a complete and coherent set of operations was performed using 922 sets of orbital baselines and 3 sets of compliant multiple orbit windows. The process strictly followed a fixed sequence: data collection, parameter matching, plotting, and synchronous export, without disrupting the order of operations. The first step involved completely collecting the original time-series information of the 922 sets of orbits and the three sets of multiple orbits. The second step involved converting the multiple orbit interval period and binding all related parameters to form a complete record. The third step involved plotting the entire orbital baseline change map, overlaying threshold lines and the three multiple orbit markers. The fourth step involved prioritizing the generation of standardized statistical tables and simultaneously exporting the corresponding analysis drawings. The completed tables and drawings can be directly used for SAR satellite multiple orbit mission planning, short baseline image selection, and quantitative analysis of orbital plane closure deviation.

[0096] The output fields, mapping specifications, and window annotation rules are standardized and defined synchronously. Short baseline candidate point marking rules are added, and a 2000m fixed threshold reference baseline drawing standard is configured, generating a standardized output configuration scheme with complete fields and matching text and graphics. All output tables are uniformly limited to four types of mandatory storage fields, without deletion or addition of custom columns. Each valid multiple orbit record fully carries four corresponding parameters, specifically: Multiple orbit time: the UTC transit time of the corresponding ascending node. Multiple orbit period: the time interval between the current window and the previous group of valid multiple orbits. Two-dimensional equivalent baseline value: the signed meter-level baseline obtained after radian conversion and equatorial radius coupling. Ascending node altitude: the equatorial transit orbit altitude of the satellite in this orbit, in kilometers. The complete data record for satellite NORAD 51284 is as follows: Orbital circle 119, critical orbit time 2025-11-23 01:31:27, critical orbit period 8 days 0 hours 0 minutes, 2D equivalent baseline -357.05m, orbital altitude 605.918km. Orbital circle 238, critical orbit time 2025-12-01 01:31:27, critical orbit period 16 days 0 hours 0 minutes, 2D equivalent baseline -455.89m, orbital altitude 605.898km. Orbital circle 357, critical orbit time 2025-12-09 01:31:24, critical orbit period 23 days 23 hours 59 minutes, 2D equivalent baseline 878.43m, orbital altitude 605.761km. The data tables generated by NORAD51822, 32376, 37216, and 39634 satellites in the same batch all use the exact same field layout and numerical storage format, with no changes in field order and no additional business columns.

[0097] The standardized drawing rules for baseline analysis diagrams are as follows: the horizontal axis of the diagram is fixedly labeled with the orbital revolution number, and the values ​​are arranged in ascending order of natural numbers from 1 to the total number of ascending nodes for the satellite; the vertical axis is uniformly labeled with the two-dimensional equivalent baseline, with the unit of measurement fixed in meters, and the vertical axis values ​​cover the range of positive and negative baseline values, adapting to the display of orbital plane offset data for all time periods. Taking the NORAD51284 satellite as an example, a total of 922 ascending nodes were extracted during the analysis period, with the horizontal axis scale ranging from 1 to 922; the vertical axis value range covers -200m to 200m, fully accommodating all baseline calculation results.

[0098] Each analysis chart includes a horizontal reference line corresponding to the baseline screening optimization threshold of 2000m. The line runs the entire vertical section of the chart without any local truncation, serving as a visual reference for intuitively distinguishing effective short baselines and false troughs. The threshold can be adjusted within the range of 1000m to 3000m based on the satellite band, and the reference line on the chart changes synchronously with the set value.

[0099] The identified compliant heavy rail points are uniformly marked with a unique symbol on the horizontal axis corresponding to the lap number, with the mark position aligned with the vertical axis value of the lap baseline. Different heavy rail windows can distinguish the mark style, making it easy to quickly identify the 1st, 2nd, 3rd and subsequent heavy rail nodes. In the NORAD51284 drawing, the laps 119, 238 and 357 are each marked with independent unique symbols, accurately corresponding to the three short baseline points.

[0100] The complete output results from NORAD satellite 51284 follow the standardized drawing and tabulation rules as follows: the horizontal axis of the drawings is labeled with orbit numbers 1 to 922, the vertical axis is in meters, and a 2000m horizontal threshold reference line is drawn throughout the entire orbit. Short baseline points are marked with special symbols for orbits 119, 238, and 357. The accompanying text tables only retain four fixed fields: re-orbit time, re-orbit period, baseline value, and orbital altitude, with no additional custom columns.

[0101] The NORAD51822, 32376, 37216, and 39634 satellite analysis charts reuse the same set of configuration standards. The coordinate axis definition, threshold lines, point markers, and table field layout are all set without differences. Multiple satellite charts can be directly compared side by side to compare the occurrence period of multiple orbits and the baseline fluctuation range, reducing the workload of batch orbit analysis and comparison.

[0102] The three rules—field standardization, threshold plotting, and point marking—are executed as a set, without being separated or adjusted individually. If the business scenario requires modification of the threshold (e.g., adjusting long-band satellites to 2800m), only the vertical coordinate of the threshold line is changed; the table fields, coordinate axis definitions, and duplicate orbit marking styles remain unchanged, ensuring a consistent format for the main output. The entire set of chart standardization rules is executed synchronously with the batch calculation of 922 sets of orbits. All valid duplicate orbit windows obtained through filtering are fully entered into the table and their corresponding points are simultaneously marked on the drawing, with no omissions in output or marking of orbit records.

[0103] This system integrates structured heavy orbit data, baseline curves of orbit cycles, and candidate window marker information, synchronously linking the ascending node altitude, baseline values, and heavy orbit period parameters for each cycle. It generates data reports and visualization analysis diagrams adapted for SAR satellite heavy orbit mission planning, short baseline image selection, and orbit closure deviation assessment. The reports use a unified fixed field layout, with each valid heavy orbit record fully encompassing four bound parameters. The field order remains fixed throughout the process, with no new custom columns added or the field order changed. The fixed field list includes orbit cycle number, heavy orbit UTC transit time, heavy orbit time interval period, two-dimensional equivalent baseline value, and ascending node altitude.

[0104] Taking the complete report of NORAD satellite 51284 as an example, orbit 119, time 2025-11-23 01:31:27, period 8 days 0 hours 0 minutes, baseline -357.05m, altitude 605.918km. Orbit 238, time 2025-12-01 01:31:27, period 16 days 0 hours 0 minutes, baseline -455.89m, altitude 605.898km. Orbit 357, time 2025-12-09 01:31:24, period 23 days 23 hours 59 minutes, baseline 878.43m, altitude 605.761km. The report is arranged in ascending order of orbit time for all compliant windows and can be directly exported and saved as a general spreadsheet file for batch indicator viewing and re-orbit cycle statistics.

[0105] The drawings employ a globally unified drawing configuration, reusing the same set of coordinate axes, baselines, and marking rules across all satellite analysis charts. The drawing includes three types of fixed elements, as follows: Continuous baseline curves across the entire orbit: The horizontal axis represents the orbital orbit number, arranged in ascending order from 1 to the total number of ascending nodes. For NORAD51284 satellite, the horizontal axis range is 1 to 922. The vertical axis unit is uniformly set to meters, plotting continuous curves of all 922 sets of two-dimensional equivalent baselines fluctuating with the orbital orbit number. 2000m horizontal threshold reference line: A horizontal straight line traversing the entire vertical range of the drawing, serving as a visual reference for distinguishing between valid and pseudo-duplicate orbits. If the screening threshold changes (within the 1000-3000m range), only the vertical coordinates of this horizontal line are adjusted; the other drawing rules remain unchanged. Short baseline-specific marker points: For each identified compliant duplicate orbit window, a unique distinguishing marker is placed at the corresponding orbital orbit on the horizontal axis and the corresponding baseline coordinates on the vertical axis. Different orders of duplicate orbits are marked with differentiated symbols.

[0106] The practical configuration of the curve chart for NORAD No. 51284 is as follows: 922 sets of baseline fluctuation curves are drawn, running through the 2000-meter horizontal baseline of the entire map. Independent markers are placed on the three horizontal axis coordinates of the track at 119, 238, and 357. The three markers correspond to the three heavy rail records in the report. The drawing and table data are completely synchronized.

[0107] Structured reports, used in conjunction with baseline plots, support three types of routine SAR orbit analysis: SAR satellite interferometric observation scheduling, used to predict image acquisition periods with short baseline conditions based on re-orbit times and intervals within the reports, and to plan ground observations and data reception. Short baseline image filtering, used in conjunction with a 2000m threshold baseline to quickly eliminate orbit data with excessively large baseline values ​​or that do not meet interferometric imaging requirements, thus narrowing the image search range. Quantitative analysis of orbital equatorial plane closure offset, used to quantify the east-west offset of orbits relative to a reference at different periods based on signed baseline values, and to analyze long-term orbital drift patterns.

[0108] Taking the NORAD51284 satellite as an example, 922 sets of full orbit baseline time series data were collected, and 3 sets of compliant re-orbit windows were filtered through a 2000m threshold. Using the orbital orbit number as an index, the baseline values, altitude, time, period, and drawing markers were globally associated and bound. Standardized tables were generated first, and the complete set of orbital indicators for orbits 119, 238, and 357 were fully entered. Baseline curves were generated according to unified drawing specifications, and 922 sets of continuous baseline curves and a 2000m threshold horizontal line were drawn. Exclusive markers were added to the three re-orbit coordinates. The tables and curves were saved as a set and directly used for engineering analysis work such as satellite re-orbit period analysis and interferometric observation mission planning.

[0109] like Figure 2 As shown, Figure 2This is a chart from a NORADID 51284 satellite re-orbit analysis report. This chart presents the baseline visualization analysis results for the SAR satellite numbered 51284. The horizontal axis represents the orbital cycle, and the vertical axis represents the longitude deviation (in degrees). The curves show the longitude shift of each orbit's ascending node relative to the reference node. Local trough points are marked, and short baseline re-orbit candidate windows meeting the 2000m threshold are highlighted with special markers. The top of the chart includes basic information such as the satellite NORAD ID, analysis start and end times, sampling parameters, total number of ascending nodes extracted, and baseline threshold. The chart also includes the orbital cycle, observation time, re-orbit period, and baseline amplitude corresponding to the first few valid re-orbits. The curves, combined with numerical annotations, visually present the plane closure deviation for each cycle, allowing for intuitive determination of the satellite's re-orbit time window and supporting interferometric data selection and mission planning.

[0110] like Figure 3 As shown, Figure 3 This is a report chart for the NORAD ID 51822 satellite multiple orbit analysis. It's a visualization of the baseline analysis specific to NORAD ID 51822 satellite. The horizontal axis represents the orbital revolutions, and the vertical axis represents the ascending node longitude deviation curve. The curve's fluctuations reflect the periodic changes in the orbital plane closure. All baseline local minimum troughs are marked within the chart, and qualified short baseline multiple orbit intervals with an absolute value ≤2000m are selected using a unique identifier. The chart header lists the satellite ID, data time range, sampling interval, number of ascending nodes extracted, and judgment threshold. It also lists the revolutions, time intervals, and baseline values ​​corresponding to multiple valid multiple orbits, visually demonstrating the satellite's multiple orbit periodicity. This chart is suitable for batch, multi-satellite unified comparative analysis.

[0111] like Figure 4 As shown, Figure 4 This is a standardized satellite baseline analysis chart for the NORAD32376 SAR satellite, used in a re-orbit analysis report. The horizontal axis represents the number of orbital orbits, and the vertical axis represents the longitude deviation, forming a two-dimensional curve that fully represents the orbital plane closure fluctuation characteristics throughout the entire time period. All local troughs are identified in the curve, distinguishing between long baseline pseudo-minimums exceeding the threshold and short baseline re-orbit windows meeting the 2000m requirement, and these distinctions are clearly marked. The chart header summarizes the basic analysis parameters, labeling the orbit number, re-orbit time, and equivalent plane baseline value for each valid re-orbit. The curve provides a visual indication of the suitable repeated observation period for interferometry of this satellite.

[0112] like Figure 5 As shown, Figure 5This is a baseline visualization report chart for NORAD ID 37216 satellite multiple orbit analysis, presented in a standardized format. The vertical axis represents the longitude deviation of the ascending node, and the horizontal axis represents the orbital revolutions. The curve depicts the long-term orbital closure trend. Local baseline troughs are automatically identified, and threshold filtering is used to retain only short-baseline multiple orbit candidate points, which are then highlighted. The chart includes the satellite ID, analysis period, baseline determination criteria, and lists the periods and baseline amplitudes of multiple orbits, clearly reflecting the variation patterns of the satellite's multiple orbit intervals and plane closure deviation.

[0113] like Figure 6 As shown, Figure 6 This is a standardized chart for NORADID 39634 satellite re-orbit analysis, adapted for NORAD 39634 satellite data. It plots baseline variation curves with orbital revolutions on the horizontal axis and longitude deviation on the vertical axis, automatically identifying all local troughs, removing invalid points with baselines exceeding 2000m, and graphically marking qualified re-orbit windows. The top of the chart uniformly labels the satellite identifier, data parameters, and core re-orbit indicators. The curves visually demonstrate the periodic closure characteristics of the satellite's orbit, allowing for quick selection of available short-baseline SAR repeat observation time series, supporting satellite observation mission planning.

[0114] In one embodiment, this application also provides a SAR satellite baseline prediction and analysis system based on TLE data, comprising: The TLE orbit data extrapolation module is used to collect the TLE orbit elements of SAR satellites, call the SGP4 orbit extrapolation algorithm to complete the orbit time series extrapolation, and generate the original nadir point time series dataset with four types of indicators: sampling time, latitude, longitude, and altitude. The time series data preprocessing module is used to perform automatic field recognition on the raw sub-satellite point dataset, standardize the time format, verify the validity of the sampled data, filter out null values ​​and abnormal distorted samples, and output a regular and continuous valid orbit time series. The ascending node interval positioning module is used to traverse the preprocessed complete time series data, retrieve adjacent sampling points whose latitude has switched from negative to non-negative values, and divide the equatorial ascending node interval corresponding to each orbit. The ascending node interpolation module is used to perform linear interpolation calculations for each ascending node interval. During the interpolation process, the longitude across ±180° meridians is simultaneously compensated and corrected with 360° compensation. The module outputs the standardized ascending node time, longitude, and altitude for each orbit. The plane baseline conversion module is used to calculate the longitude closure error of the entire circle based on the longitude of the ascending node of the first circle and normalize it to the range of [-180°, 180°]. Combined with the Earth's equatorial radius, it completes the angle to meter-level calculation to obtain the two-dimensional sub-satellite point equivalent plane baseline. The short baseline window filtering module is used to traverse the entire orbital baseline sequence to identify local minimum valleys, and uses a fixed threshold of 2000m to filter candidate circles to select effective short baseline re-orbit windows that meet the interference requirements. The re-track results output module is used to collect baseline values ​​and all information from the filtering window, standardize the output of re-track time, period, and orbital height structured data, and simultaneously generate a visualization analysis curve of orbital cycles and baseline changes.

[0115] The computer-readable storage medium provided in the above embodiments of this application and the SAR satellite baseline prediction analysis method based on TLE data provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application stored therein.

[0116] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for evaluating a SAR satellite baseline prediction analysis method, system, electronic device, and readable storage medium based on TLE data are basically similar to the aforementioned embodiment of a SAR satellite baseline prediction analysis method based on TLE data, and therefore the description is relatively simple. Relevant parts can be referred to in the description of the aforementioned embodiment of a SAR satellite baseline prediction analysis method based on TLE data.

Claims

1. A SAR satellite baseline prediction and analysis method based on TLE data, characterized in that, include: Acquire SAR satellite TLE orbital element data, call the SGP4 orbital extrapolation module to complete orbital extrapolation, and generate a nadir point time series dataset containing time, latitude, longitude, and altitude; perform data column identification on the dataset, complete time standardization and data validity verification, and remove abnormal and null sampled data; Traverse the verified nadir point time series data, identify the zero-crossing intervals where latitude transitions from negative to positive values, and lock the ascending node interval corresponding to each orbit. Linear interpolation is performed for each ascending node interval, with longitude correction performed simultaneously during the interpolation process, to obtain the time, longitude, and altitude parameters corresponding to the ascending node of each orbital cycle; Using the longitude of the ascending node of the first circle as the benchmark reference value, the longitude closure error of the remaining ascending nodes is calculated for each circle, and the error range is normalized to the continuous interval of [-180°, 180°]. By combining the Earth's equatorial radius parameter, the normalized ascending node longitude closure error is converted into a meter-level two-dimensional sub-satellite point equivalent plane baseline; Local troughs were identified in the baseline numerical sequence of all orbital cycles to screen out candidate re-orbit cycles corresponding to local minimum values ​​of the baselines; a fixed threshold of 2000m was set to filter the candidate re-orbit cycles, and samples with baseline values ​​not exceeding the threshold were retained to obtain an effective short baseline re-orbit window; The equivalent plane baseline calculation results are integrated with the filtered short baseline re-rail window information to uniformly output the re-rail time, re-rail cycle, baseline value, and track height data, and simultaneously generate a track cycle-baseline change analysis curve.

2. The method as described in claim 1, characterized in that, Linear interpolation is performed for each ascending node interval, with longitude correction performed simultaneously during the interpolation process. This yields the time, longitude, and altitude parameters corresponding to the ascending node of each orbital cycle, including: For the sampling points of the ascending node interval where the latitude changes from negative to positive, interval locking and numerical analysis are performed to extract the original sampling data of the first and last time of the interval, latitude, longitude and altitude, and to construct the interpolation base dataset and the longitude discrimination logic across ±180° meridians; The linear variation ratio of latitude in the interval, the 360° longitude pre-compensation rule, and the standardization of the interpolation domain are standardized to generate linear interpolation calculation rules and longitude circumference correction execution details for the adaptation of orbit ascending node solution. The interpolation dataset, linear interpolation calculation rules, and longitude correction execution details are integrated. Based on the orbital orbit ascending node solution logic, cross-boundary longitude pre-compensation correction, linear interpolation calculation, and result range normalization are completed to generate standardized ascending node time, longitude, and altitude parameters for each orbital orbit. The linear interpolation calculation rules include... ; ; ; ; ,in, Let be the latitude of the satellite nadir point corresponding to the i-th orbital timing sampling point; The latitude of the satellite nadir point corresponding to the (i+1)th adjacent orbit time-series sampling point; The linear interpolation weights are for the latitudinal dimensions. This is the observation timestamp corresponding to the i-th sampling point; This is the observation timestamp corresponding to the (i+1)th adjacent sampling point; The precise time of the satellite's ascending node obtained by linear interpolation (the standard time when the satellite's latitude is equal to 0 and it crosses the equator). The original longitude of the nadir point corresponding to the i-th sampling point; The original longitude of the sub-satellite point corresponding to the (i+1)th adjacent sampling point; The standardized ascending node longitude obtained through interpolation; The altitude of the satellite orbit corresponding to the i-th sampling point; The satellite orbital altitude corresponding to the (i+1)th adjacent sampling point; The altitude of the satellite's ascending node is obtained through interpolation.

3. The method as described in claim 1, characterized in that, Using the longitude of the ascending node of the first orbit as the reference value, the longitude closure error of the remaining ascending nodes is calculated for each orbit, and the error range is normalized to the continuous interval [-180°, 180°], including: For the longitude dataset of the ascending node of all orbital cycles, benchmark assignment and closure error modeling are carried out, adopting a three-step progressive architecture of benchmark longitude selection, original error solution for each cycle, and normalization of out-of-limit value range. The baseline anchoring was completed according to the order of the orbital orbits, and the longitude of the ascending node of the first orbit was determined. Set the global baseline reference value, divide the baseline longitude dataset and the ascending node longitude dataset of each circle to be calculated, and clarify the calculation boundaries and input purposes of the two types of data; For the two sets of longitude datasets, a standardized closure error calculation logic was built, first according to the formula. The original longitude closure error is solved circle by circle, and then ±360° compensation correction is performed on the error that exceeds the range of [-180°, 180°]. like ;like This completes the output of the normalized longitude closure error.

4. The method as described in claim 1, characterized in that, Based on the Earth's equatorial radius parameter, the normalized ascending node longitude closure error is converted into a meter-level two-dimensional sub-satellite equivalent plane baseline, including: Baseline dimension transformation modeling was carried out on the normalized longitude closure error dataset, using a three-step progressive architecture of radian conversion, radius coupling, and plane baseline solution. Based on the mathematical rules of angle-length conversion, parameter matching is implemented, and the normalized longitude closure error and the Earth's equatorial radius are divided into an angle input subset and a length reference subset, clarifying the calculation boundaries and unit matching requirements of the two types of parameters; The equivalent baseline conversion calculation logic is built for the two types of parameters, according to the formula. Complete the angle-radius conversion and equatorial radius coupling calculation, and output the meter-level two-dimensional sub-satellite point equivalent plane baseline.

5. The method as described in claim 1, characterized in that, Local troughs were identified in the baseline numerical sequence of all orbital cycles to screen candidate re-orbit cycles corresponding to local minimum baseline values. A fixed threshold of 2000m was set to filter the candidate re-orbit cycles, retaining samples whose baseline values ​​did not exceed the threshold, thus obtaining an effective short baseline re-orbit window, including: Combining the operational requirements of SAR short baseline screening with the logic of baseline sequence extreme value discrimination, a mechanism for determining local valleys, fixed threshold filtering, and candidate window extraction is introduced to determine a standardized processing scheme for baseline sequence traversal, minimum value identification, threshold screening, and duplicate orbit window output. A global traversal and local trough discrimination are performed on the two-dimensional equivalent baseline sequence of the entire orbital cycle to extract the data that meet the requirements. The absolute values ​​of the baselines of adjacent preceding and following orbits are both greater than the local minimum candidate orbit number of the current orbit. This is the two-dimensional equivalent baseline corresponding to the k-th orbit; The absolute value of the equivalent baseline in the kth circumference represents only the magnitude of the offset distance. The current orbit is the orbit immediately preceding the previous orbit. This is the absolute value of the baseline of the previous circumference; It is the two-dimensional equivalent baseline of the next adjacent orbit after the current orbit (the (k+1)th orbit); This represents the absolute value of the baseline for the next lap. A 2000m baseline screening threshold system was constructed based on SAR interferometry optimization constraints to distinguish between two types of sample boundaries: invalid loops exceeding the threshold and effective short base loops within the threshold, satisfying the following conditions: The condition is that the compliant re-railway window extraction is completed.

6. The method as described in claim 5, characterized in that, Integrating the equivalent plane baseline calculation results with the filtered short baseline re-rail window information, the system uniformly outputs re-rail time, re-rail period, baseline values, and track altitude data, and simultaneously generates a track cycle-baseline change analysis curve, including: The system adopts an integrated approach that combines baseline statistical collection, duplicate track information fusion, visualization plotting, and simultaneous output of multiple results. It unifies the output process, matches the granularity of baseline data collection according to the time sequence of the track circle, sets the execution order of structured numerical output and baseline trend chart drawing according to business priorities, and divides the output into two types of output carriers: standardized data tables and baseline analysis charts. The output fields, mapping specifications, and window annotation rules are defined in a standardized manner. Short baseline candidate point marking rules are added, and a 2000m fixed threshold reference baseline drawing standard is configured to generate a standardized output configuration scheme with complete fields and matching text and graphics. By integrating structured heavy orbit data, baseline curves of orbital loops, and candidate window marking information, and synchronously linking the ascending node altitude, baseline values, and heavy orbit period parameters of each loop, data reports and visualization analysis drawings adapted to SAR satellite heavy orbit mission planning, short baseline image selection, and orbit closure deviation assessment are generated.

7. A SAR satellite baseline prediction and analysis system based on TLE data, characterized in that, The system is used to execute executable instructions to implement the SAR satellite baseline prediction and analysis method based on TLE data as described in any one of claims 1 to 6.

8. An electronic device, characterized in that, include: First processor; The processor also includes a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the SAR satellite baseline prediction analysis method based on TLE data according to any one of claims 1 to 6 by executing the executable instructions.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the SAR satellite baseline prediction and analysis method based on TLE data as described in any one of claims 1 to 6.

Images