Dual-star cooperative multi-type abnormal spectrum information fusion method and system

Abstract

The application discloses a dual-satellite cooperative multi-type abnormal air spectrum information fusion method and system, relates to the technical field of dual-satellite cooperation, and comprises the following steps: in the starting stage of dual-satellite cooperative observation, a cross-satellite unified time baseline is established, nanosecond-level alignment of time series observed by two satellites is realized through a high-precision inter-satellite communication link, and an orbit perturbation parameter caused by orbit micro-perturbation is injected into the unified time baseline, and a drift prediction matrix with dynamic prediction capability is generated based on the orbit perturbation parameter; under the constraint condition of the drift prediction matrix. The application realizes nanosecond-level time alignment of dual-satellite observation through the cross-satellite unified time baseline and the drift prediction matrix, improves the phase consistency and energy stability of air spectrum fusion, and realizes real-time correction and stable regulation of the fusion result through a phase conjugate compensation chain and a counterfactual playback mechanism to inhibit false enhancement, thereby improving the precision and reliability of ground surface anomaly monitoring.

Description

Technical Field

[0001] This invention relates to the field of binary star synergy technology, specifically to a method and system for fusing multi-type anomaly spatial spectrum information in binary star synergy. Background Technology

[0002] The dual-satellite collaborative fusion of multi-type anomaly spatial-spectral information mainly refers to the acquisition of multi-dimensional remote sensing data of the Earth's surface region by two orbiting satellites, one equipped with a multispectral camera (Mudu-1A) and the other with a hyperspectral camera (Mudu-1B), at the same or similar time points. The spectral information is then complemented and fused through on-board or satellite-ground joint processing. Its core lies in leveraging the wide coverage and high timeliness of multispectral data to quickly locate and mark potentially existing anomaly areas on the Earth's surface; combined with the fine-grained spectral resolution of hyperspectral data, precise diagnosis and attribute analysis of the target area are performed. Through this dual-satellite collaborative observation and information fusion mechanism, the spatial distribution characteristics and spectral response characteristics of surface anomalies can be synchronously integrated to construct a spatiotemporally integrated anomaly information map. Furthermore, spatial-spectral fusion is not limited to simple overlay; it involves radiometric correction, geometric correction, feature extraction, and deep learning model recognition of multi-source data through on-orbit processors and integrated processing units, thereby enhancing anomaly signals and suppressing false signals. This method effectively enhances the rapid identification capability of sudden and dynamic surface anomalies (such as subsidence, landslides, thermal anomalies, and water pollution) and reduces the problem of insufficient spatiotemporal or spectral information that may result from a single sensor. In other words, the fusion of multi-type anomaly spatial and spectral information through dual-satellite collaboration complements the macroscopic coverage of 1A with the fine analysis of 1B. Through collaborative modeling and fusion processing of multi-dimensional information, a more accurate and timely surface anomaly monitoring and early warning system is formed, providing support for disaster prevention and control, resource and environmental monitoring, and emergency response.

[0003] The existing technology has the following shortcomings: During the collaborative observation of asynchronous imaging of two satellites, the spectral acquisition window may experience millisecond-level temporal misalignment due to the dynamic effects of minute orbital perturbations. This misalignment can trigger anomalous phase jumps in the spatial-spectral fusion process, leading to a large-scale false enhancement effect in the fused surface energy distribution map. Once such anomalies occur, actual local surface anomalies will be misjudged as large-scale energy surges, severely interfering with the reliability of surface anomaly monitoring results and potentially causing significant deviations in disaster identification and early warning.

[0004] The information disclosed in the background section is only intended 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

[0005] The purpose of this invention is to provide a method and system for fusing multi-type anomalous spatial spectrum information in binary star collaboration, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for fusing multi-type anomaly spatial spectrum information through binary star collaboration, comprising the following steps: In the initial stage of dual-satellite collaborative observation, a unified time baseline is established across satellites. Nanosecond-level alignment of the observation time series of the two satellites is achieved through a high-precision inter-satellite communication link. Orbital perturbation parameters caused by minor orbital disturbances are injected into the unified time baseline. A drift prediction matrix with dynamic prediction capabilities is generated based on the orbital perturbation parameters to serve as the initial basis for subsequent spectral acquisition time correction. Under the constraint of the drift prediction matrix, time-slip calculation is performed on the spectral acquisition windows of the two satellites to extract the potential time misalignment intervals caused by small orbital perturbations. A time-series anchor group is generated based on the time misalignment intervals, which is used to provide a synchronization reference for subsequent spectral phase compensation. A phase conjugate compensation chain is constructed based on the time-series anchor group to invert and model the spectral phase anomaly caused by the time misalignment interval. The phase compensation signal is dynamically injected into the spatial-spectral fusion channel so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process, thereby suppressing the false enhancement effect of energy distribution caused by the abnormal phase. After the phase conjugate compensation chain completes real-time phase suppression, the counterfactual playback mechanism based on synchronization reference is invoked to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval. The residual false enhancement signals in the fusion result are identified and eliminated, and the differential comparison result is written back to the unified time baseline to form a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation.

[0007] Preferably, the steps for establishing a unified time baseline across satellites include: First, the ground control center issues a time synchronization command, establishes a two-way time calibration channel through a high-precision inter-satellite communication link between satellites, and uses the round-trip ranging delay difference for correction to achieve nanosecond-level alignment of the observation time series of the two satellites; Second, the orbital perturbation parameters caused by minor orbital disturbances are injected into the unified time baseline to give it dynamic response characteristics; Then, a drift prediction matrix with dynamic prediction capabilities is generated based on the orbital perturbation parameters, and weighted fitting and recursive prediction are performed on each perturbation component to obtain an estimate of the time drift; Finally, during the spectral acquisition stage, the spectral acquisition trigger time is slipped based on the prediction results in the drift prediction matrix, and closed-loop time correction is achieved through the round-trip delay feedback of the inter-satellite link, thereby maintaining the time consistency of the two satellites.

[0008] Preferably, in the process of generating the drift prediction matrix, the weighted fitting of the orbital perturbation parameters adopts a combination of Kalman filtering and polynomial extrapolation to achieve continuous prediction of the time drift. During satellite operation, the internal weight coefficients are automatically adjusted according to the real-time feedback of the inter-satellite link round-trip delay, thereby improving the prediction accuracy and dynamic stability of the drift prediction matrix for the time offset caused by orbital perturbation.

[0009] Preferably, the steps for performing time-slip calculation under the constraints of the drift prediction matrix include: First, jointly matching the drift prediction matrix with the on-orbit attitude information, imaging time stamps, and ground reference time series of the two satellites, dynamically correcting the observation time, and calculating the initial estimate of the time offset; Second, introducing a weighted least squares slip fitting algorithm to approximate the time series piecewise and impose continuous constraints to obtain the time-slip calculation result; Then, using the drift sensitivity parameter of the drift prediction matrix as a constraint, extracting the potential time misalignment intervals caused by orbital disturbances and forming a time misalignment parameter set; Finally, generating a time-series anchor group based on the time misalignment parameter set, and achieving fine-tuning and dynamic correction of the anchor time through coupling verification with the drift prediction matrix and the time-slip calculation result to ensure the temporal continuity and spatial consistency of the time-series anchor group.

[0010] Preferably, in the step of generating the time-series anchor group, the time coordinates of each anchor point are compared with the predicted drift values ​​in the drift prediction matrix. When the deviation exceeds the allowable threshold, the anchor point time is fine-tuned based on the time-slip solution results, and real-time correction is performed through the dynamic prediction model of the drift prediction matrix to ensure the stability and reliability of the time-series anchor group in terms of temporal continuity and spatial correspondence.

[0011] Preferably, the steps for constructing a phase conjugate compensation chain based on a time-series anchor group include: First, aligning the time-series anchor group with the spectral acquisition data of two satellites in the same surface area in the time domain, and calculating the spectral phase shift function based on the time weight and spatial coordinates of each anchor point; Second, extracting the phase shift component based on the spectral phase shift function and performing spectral decomposition, establishing inversion equations for the low-frequency trend term, high-frequency disturbance term, and transient phase change term respectively, and generating a continuous phase conjugate response function; Then, embedding the phase conjugate response function into the spatial-spectral fusion channel to achieve dynamic injection of the phase compensation signal, and achieving real-time spectral phase correction through adaptive weight allocation of land cover categories; Finally, after compensation is completed, using the sliding observation window of the time-series anchor group for real-time monitoring, and correcting the weight of the phase conjugate response function through the drift prediction matrix to achieve closed-loop steady-state optimization of the compensation chain.

[0012] Preferably, in the process of embedding the phase conjugate response function into the spatial-spectral fusion channel and realizing the dynamic injection of the phase compensation signal, a phase energy constraint mechanism is further introduced to normalize and restrict the spectral energy distribution after fusion, and the weight of the phase conjugate response function is adaptively optimized by combining the time domain differential feedback mechanism to prevent excessive energy enhancement or phase drift accumulation during the compensation process, and to ensure the steady-state consistency of the compensated spectral signal in the time domain and energy domain.

[0013] Preferably, the steps of invoking the counterfactual playback mechanism based on synchronization reference include: First, establishing a playback time window for the phase-compensated fused data using the synchronization reference time series defined by the time-series anchor group, and calculating the phase backtracking difference of the spectral signals before and after compensation to form a phase playback trajectory; Second, performing differential comparison between the phase-compensated spatial-spectral fusion result and the historical trajectory data corresponding to the time misalignment interval to identify and mark the residual false enhancement signals; Then, performing energy backtracking operation on the false enhancement signals based on the differential comparison results, and achieving rematching and stable recovery of the spectral signals through a counterfactual probability correction strategy; Finally, integrating the backtracked spectral energy distribution and phase residual information to generate a time-series error correction vector, and writing the correction vector back to the unified time baseline to achieve closed-loop steady-state time synchronization and energy correction.

[0014] Preferably, in the energy backoff operation, the phase conjugate inverse transform is used to subtract the conjugate value of the phase shift from the fused signal, and after backoff, time-domain smoothing is used to eliminate local non-physical oscillations. At the same time, a counterfactual probability correction threshold is set based on the energy fluctuation characteristics of similar land features in historical trajectories to limit the energy backoff amplitude and prevent the true spectral signal from being excessively weakened.

[0015] A dual-star collaborative multi-type anomaly spatial spectrum information fusion system includes a time baseline construction module, a time-slip calculation module, a phase conjugate compensation module, and a counterfactual playback correction module; The time baseline construction module establishes a unified time baseline across satellites at the beginning of the dual-satellite collaborative observation. It achieves nanosecond-level alignment of the observation time series of the two satellites through a high-precision inter-satellite communication link, and injects the orbital perturbation parameters caused by minor orbital disturbances into the unified time baseline. Based on the orbital perturbation parameters, it generates a drift prediction matrix with dynamic prediction capabilities. The time-slip solution module performs time-slip solution on the spectral acquisition windows of the two satellites under the constraint of the drift prediction matrix, extracts the potential time misalignment intervals caused by small orbital perturbations, and generates a time-series anchor point group based on the time misalignment intervals; The phase conjugate compensation module constructs a phase conjugate compensation chain based on the time-series anchor group, performs inversion modeling of spectral phase anomalies caused by time misalignment intervals, and dynamically injects phase compensation signals into the spatial-spectral fusion channel, so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process. The counterfactual playback correction module, after the phase conjugate compensation chain completes real-time phase suppression, calls the counterfactual playback mechanism based on synchronization reference to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval, identify and remove residual false enhancement signals in the fusion result, and write the differential comparison result back to the unified time baseline, forming a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention establishes a unified cross-satellite time baseline at the initial stage of dual-satellite collaborative observation and generates a drift prediction matrix with dynamic prediction capabilities by combining orbital perturbation parameters. This achieves nanosecond-level time alignment and dynamic drift correction during the spectral acquisition process of the two satellites. This method maintains continuous time consistency between the two satellites throughout the observation mission, effectively eliminating the impact of orbital perturbations and on-board clock drift on the synchronization accuracy of the spectral acquisition window. This ensures strict temporal correspondence between the acquired data from both satellites, allowing the subsequent spatial-spectral fusion process to be based on high-precision time consistency. This time unification mechanism significantly improves the phase consistency and energy superposition stability of multi-source data during fusion, avoiding the spread of fusion errors caused by time offsets.

[0017] This invention achieves real-time suppression and closed-loop correction of spectral phase anomalies caused by temporal misalignment during spatial-spectral fusion by constructing a phase conjugate compensation chain based on a time-series anchor group and a counterfactual playback mechanism. This technical approach enables dynamic injection of phase compensation signals and differential comparison correction during the fusion process, allowing residual spurious enhancement signals to be identified and eliminated at the generation stage, thereby significantly improving the physical authenticity and energy distribution stability of the fusion results. This closed-loop time synchronization and energy correction mechanism not only effectively suppresses spurious enhancement effects caused by orbital perturbations but also enables the dual-satellite collaborative observation system to possess adaptive steady-state control capabilities, significantly improving the accuracy and reliability of surface anomaly monitoring, disaster identification, and environmental early warning. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart of the method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to the present invention.

[0020] Figure 2 This is a schematic diagram of the modules of the dual-star collaborative multi-type anomaly spatial spectrum information fusion system of the present invention. Detailed Implementation

[0021] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0022] This invention provides, for example Figure 1 The binary star-coordinated multi-type anomaly spatial spectrum information fusion method shown includes the following steps: In the initial stage of dual-satellite collaborative observation, a unified time baseline is established across satellites. Nanosecond-level alignment of the observation time series of the two satellites is achieved through a high-precision inter-satellite communication link. Orbital perturbation parameters caused by minor orbital disturbances are injected into the unified time baseline. A drift prediction matrix with dynamic prediction capabilities is generated based on the orbital perturbation parameters to serve as the initial basis for subsequent spectral acquisition time correction. The specific implementation steps are as follows: In the initial phase of dual-satellite collaborative observation, to ensure high-precision time consistency between the two orbiting satellites during data acquisition, a unified cross-satellite time baseline needs to be established. To achieve this, the ground control center first issues a time synchronization command, establishing a mutually visible two-way time calibration channel through a high-precision inter-satellite communication link. During this process, the system sends and receives highly stable time pulse signals via the inter-satellite link and uses the difference in two-way ranging delay for correction, transforming the time references of the two satellites from their original independent local clocks into a shared unified time reference. To achieve nanosecond-level time alignment accuracy, a time-frequency comparison method based on atomic clock reference frequencies is employed during synchronization. By measuring the round-trip time delay between satellites and performing multiple averaging processes, minute deviations caused by changes in spatial distance and electromagnetic wave propagation medium are corrected, thus forming a unified cross-satellite time baseline. This unified time baseline not only provides an absolute reference for the synchronization moment but also retains the statistical mapping relationship between time offset and propagation path changes, providing a traceable time reference for subsequent dynamic orbit compensation.

[0023] After establishing a unified time baseline, orbital perturbation parameters are injected into it to further overcome the dynamic impact of minor orbital perturbations on time synchronization. These parameters include long-term perturbation terms caused by Earth's non-spherical gravitational field, periodic drift terms caused by atmospheric drag, short-period perturbation terms caused by solar radiation pressure, and instantaneous velocity deviation terms caused by attitude adjustments. To achieve precise injection, the ground-based orbit control center first performs real-time inversion calculations on the state vector of each satellite based on the latest orbital elements, generating orbital perturbation components. These perturbation components are then formatted into time function expressions and transmitted to the two satellites via inter-satellite communication links. Upon receiving the data, each satellite maps the time series of these orbital perturbation parameters onto the time scale of the unified time baseline, enabling orbital dynamics changes to reflect predictable drift trends on the time baseline. In this way, the unified time baseline evolves from a static synchronization framework into a time reference system with dynamic response characteristics, maintaining time consistency even with minor changes in the satellite's trajectory, thus providing a solid foundation for subsequent drift modeling and spectral acquisition time correction.

[0024] After injecting orbital perturbation parameters, a drift prediction matrix with dynamic predictive capabilities is generated based on these parameters. To construct this matrix, firstly, a weighted fit is performed on each perturbation component according to the time series of the orbital perturbation parameters to establish its sensitivity function to time shift. Then, using a combination of Kalman filtering and polynomial extrapolation, the trend of time shift with orbital position is recursively predicted to obtain the drift estimate for each moment. Next, all estimates are filled into the two-dimensional index space of the prediction matrix using time and orbital position as indices, forming a dynamic matrix that can reflect the relationship between orbital perturbations and time drift in real time. This drift prediction matrix not only records the direction and magnitude of drift at different time points but also has an adaptive capability to be automatically updated based on observation data. As the two satellites continue to operate along their orbits and receive new inter-satellite time-frequency alignment data, the drift prediction matrix automatically adjusts its internal weight coefficients according to the latest time shift feedback, thereby maintaining its forward-looking predictive capability for the drift effect caused by orbital perturbations. In this way, the drift prediction matrix becomes a dynamic bridge connecting orbital perturbations and time offsets, providing a calculable predictive basis for the accurate correction of subsequent spectral acquisition time.

[0025] After obtaining the drift prediction matrix, it is used as the initial basis for subsequent spectral acquisition time correction. During the spectral data acquisition phase, the two satellites initiate their respective spectral acquisition programs. However, due to the dynamic influence of perturbations, there may still be residual time misalignments in the spectral acquisition window, ranging from microseconds to milliseconds. At this time, the dynamic prediction results stored in the aforementioned drift prediction matrix can be used to slide and correct the trigger time of spectral acquisition. Specifically, before each acquisition, the satellite extracts the corresponding drift amount from the drift prediction matrix based on its current orbital position and time index, and adds it as a time offset correction value to the original acquisition trigger signal, thereby achieving dynamic time compensation for the spectral acquisition window. Simultaneously, to prevent over-correction or delay accumulation during the compensation process, the system also monitors the round-trip delay changes of the inter-satellite link in real time and feeds this information back to the adaptive update mechanism of the drift prediction matrix, forming a closed-loop iteration between drift prediction and time correction. After this closed-loop dynamic correction, the time error of the dual-satellite spectral acquisition can be stably controlled within the nanosecond range, ensuring that the spectral sampling points of the two satellites in collaborative observation strictly correspond to the surface response under the same time baseline.

[0026] Through this continuous process of cross-satellite unified time baseline, orbital perturbation parameter injection, drift prediction matrix generation, and time slip correction, a closed-loop control system from time synchronization and perturbation modeling to dynamic compensation is achieved. This completely eliminates the interference of time misalignment caused by orbital perturbations on the accuracy of spatial-spectral fusion, laying a stable and reliable time consistency foundation for subsequent phase compensation and energy fusion in dual-satellite collaborative observation.

[0027] Under the constraint of the drift prediction matrix, time-slip calculation is performed on the spectral acquisition windows of the two satellites to extract the potential time misalignment intervals caused by small orbital perturbations. A time-series anchor group is generated based on the time misalignment intervals, which is used to provide a synchronization reference for subsequent spectral phase compensation. The specific implementation steps are as follows: To achieve time-slip calculation of the spectral acquisition windows of two satellites under the constraints of the drift prediction matrix, the drift prediction matrix generated in the previous step is first jointly matched with the on-orbit attitude information, imaging time markers, and ground reference time series of the two satellites. Specifically, firstly, based on the dynamic time drift recorded in the drift prediction matrix, the observation times of the two satellites within the current orbital period are dynamically corrected. The corrected time series is then mapped one-to-one with the absolute time baseline provided by the ground control center to form a standardized time calculation input set. Next, the initial estimate of the time offset is calculated by comparing the differences in spectral sampling times of the two satellites within the same observation area. To eliminate nonlinear time drift caused by electromagnetic propagation delay, attitude fine-tuning, or data transmission time differences, this step further introduces a weighted least squares-based slip fitting algorithm to approximate the time series piecewise and impose continuous constraints, ensuring that the offset at each moment conforms to the dynamic prediction trend provided by the drift prediction matrix while adaptively correcting local random errors. After multiple rounds of iterative calculations, preliminary results of the time slip between the spectral acquisition windows of the two satellites can be obtained, providing high-precision basic data for subsequent time misalignment extraction.

[0028] After obtaining the time-slip calculation results, potential time misalignment intervals caused by minor orbital perturbations are extracted. This process uses the drift sensitivity parameters provided by the drift prediction matrix as constraints and the time-slip calculation results as input data to identify the boundaries of time misalignment intervals through time difference analysis. Specifically, firstly, the time difference curves for consecutive observation times are calculated based on the results of the slip calculation. When the rate of change of the time difference curve exceeds the stability threshold defined by the drift prediction matrix, it is determined that there is a potential time misalignment in the current interval. To further improve the accuracy of the extraction, cross-validation is performed using the orbital attitude perturbation function and the spectral acquisition time series function to ensure that the identified time misalignment intervals do indeed originate from orbital perturbations rather than other non-systematic noise. For each detected time misalignment interval, a weighted time series smoothing algorithm is used to calculate its center drift, duration, and phase shift direction, thus forming a complete set of time misalignment parameters. This parameter set reflects both the time boundary of the misalignment and quantifies its actual impact on the synchronization of spectral acquisition, providing a quantitative basis for the subsequent establishment of a time series anchor group.

[0029] After acquiring the time misalignment parameter set, a time-series anchor group is generated based on these parameters to provide a reliable synchronization reference for spectral phase compensation. The generation process of the time-series anchor group essentially transforms discrete time misalignment information into a continuous reference system with temporal and spatial correspondences. First, based on the center drift amount in the time misalignment parameter set, the key moments of each misalignment interval are determined as the initial anchor positions. Then, by comparing the sampling trajectories of two satellites in the same surface area, the spatial correspondence of each anchor point is calculated and projected onto a unified time baseline. Next, using the dynamic prediction model in the drift prediction matrix, the time coordinates of these anchor points are fine-tuned so that each anchor point reflects the dynamic changes of orbital perturbations while maintaining consistency with the time-slip solution results. To enhance the stability of the time-series anchor group, a time-weighted constraint strategy is introduced to hierarchically sort anchor points with different weights, giving higher reference priority to anchor points closer to the center of the spectral acquisition window, while anchor points at the window edges serve as auxiliary alignment points. Ultimately, all the weighted and adjusted anchor points together form a continuous and updatable time-series anchor point group, covering the entire observation period and reflecting the dynamic time shift characteristics caused by orbital disturbances.

[0030] After generating the time-series anchor group, to ensure its effective role as a synchronization reference in subsequent spectral phase compensation, the time-series anchor group is coupled and verified with the drift prediction matrix and time-slip calculation results. Specifically, firstly, the time coordinate of each anchor point is compared item by item with the predicted drift value in the drift prediction matrix. If the deviation exceeds the allowable threshold, the anchor point time is slightly adjusted according to the slip calculation results until the deviation converges within the allowable range. Subsequently, the adjusted anchor point time series is interpolated and matched with the spectral sampling time series of the two satellites to ensure that there are one or more valid anchor points in each effective sampling window of spectral acquisition for calibrating the synchronization reference time. To prevent anchor point drift accumulation due to abrupt changes in orbital disturbances, the temporal continuity and spatial consistency of the anchor group need to be monitored in real time. When an abnormal drift trend is detected, the anchor point position is immediately corrected by calling the dynamic prediction model of the drift prediction matrix, thereby ensuring the stability of the entire time-series anchor group in both time and spatial dimensions.

[0031] Through the above process, the resulting time-series anchor point group can not only accurately reflect the time misalignment characteristics caused by orbital disturbances, but also provide a stable and highly confident synchronization reference for each moment in the spectral phase compensation stage, enabling the data from different satellites to be accurately aligned in the time domain during the subsequent spatial-spectral fusion process, significantly improving the reliability and timeliness of surface anomaly information fusion.

[0032] A phase conjugate compensation chain is constructed based on the time-series anchor group to invert and model the spectral phase anomaly caused by the time misalignment interval. The phase compensation signal is dynamically injected into the spatial-spectral fusion channel so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process, thereby suppressing the false enhancement effect of energy distribution caused by the abnormal phase. The specific implementation steps are as follows: In constructing a phase conjugate compensation chain based on a time-series anchor group, the time-series anchor group generated in the previous step is first time-domain aligned with the spectral acquisition data of the two satellites in the same surface area to form a basic reference set for phase inversion. Specifically, the time coordinate of each anchor point in the time-series anchor group is first matched with the corresponding multispectral data sampling time and hyperspectral data sampling time to establish a one-to-one time index mapping. Then, based on the time weight and spatial coordinate of each anchor point, the spectral response curves of the two satellites under the same ground target are synchronized and normalized to obtain a spectral signal sequence under a unified time baseline. On this basis, by comparing the relative phase shift of the spectral curves before and after synchronization, the spectral phase deviation distribution function caused by the time misalignment interval is calculated. To ensure the accuracy of the calculation results, this process adopts a multi-level fitting strategy: first, the spectral signal is piecewise corrected using the time reference provided by the time-series anchor group; then, a nonlinear least squares method is introduced to smoothly fit the phase shift curve, thereby obtaining a continuous and differentiable phase shift function. This phase shift function reveals the intrinsic correspondence between the time misalignment caused by orbital perturbations and the spectral phase anomaly, providing an accurate input basis for subsequent inversion modeling.

[0033] After obtaining the spectral phase shift function, an inversion model is performed to model the phase anomalies caused by time misalignment intervals, establishing a dynamic solution mechanism for phase conjugate compensation. Specifically, firstly, the phase shift components within each misalignment interval are extracted based on the phase shift function and their spectroscopic components are decomposed into three categories: low-frequency trend terms, high-frequency disturbance terms, and transient phase change terms. Then, corresponding inversion equations are established for different types of phase shift characteristics. For the low-frequency trend term, a linear phase drift model is used for inversion; for the high-frequency disturbance term, an inverse Fourier transform is introduced to establish a phase recovery model from the frequency domain to the time domain; and for the transient change term, a probabilistic inversion method based on Bayesian estimation is used to perform smooth compensation using information from neighboring anchor points in the time-series anchor point group. Each inversion model uses the time-series anchor point group as a synchronous reference in the time domain to ensure consistency between the inversion calculation and the time drift prediction. After multidimensional inversion calculation, the phase compensation value corresponding to each time step is obtained, forming a continuous phase conjugate response function. The significance of this function lies in the fact that it is a mirror image of the phase offset function in time, and mathematically realizes the reverse mapping of phase error, providing theoretical support for constructing a phase conjugate compensation chain.

[0034] After obtaining the phase conjugate response function, it is embedded into the spatial-spectral fusion channel to achieve dynamic injection of the phase compensation signal, thereby suppressing phase anomalies in real time during data fusion. Specifically, firstly, based on the synchronization reference determined by the time-series anchor group, each time node in the phase conjugate response function is aligned with the pixel during the fusion of multispectral and hyperspectral data. In each pixel fusion operation, the current compensation value of the phase conjugate response function is superimposed onto the complex representation of the hyperspectral signal, thus achieving instantaneous correction of the spectral phase. Simultaneously, considering the differences in phase sensitivity of different land cover reflectance characteristics, an adaptive weight allocation mechanism based on land cover category is introduced during phase compensation injection, setting different compensation coefficients for different surface areas to avoid spectral distortion caused by over-correction. Within the fusion channel, all corrected spectral signals achieve phase consistency in the time domain, thus forming a coherent superposition effect during the spectral fusion stage, significantly suppressing spurious enhancement of energy distribution caused by temporal misalignment. To ensure real-time performance, the compensation injection process is performed simultaneously with data fusion, and the time-series anchor group is updated with synchronous references in each fusion cycle to ensure the dynamic consistency and fusion stability of the phase conjugate compensation signal.

[0035] After the dynamic injection of the phase conjugate compensation chain is completed, the compensation effect is enhanced and optimized in real time to ensure that the entire spatial-spectral fusion process reaches a steady-state balance in both time and energy dimensions. Specifically, firstly, the continuous time period defined in the time-series anchor group is used as a sliding observation window to monitor the spectral phase consistency within each fusion cycle. When a phase drift trend is detected, the dynamic prediction result in the drift prediction matrix is ​​immediately invoked to correct the weights of the phase conjugate response function, ensuring that its compensation direction is opposite to the drift trend. Secondly, to prevent noise accumulation during the compensation process, a phase energy constraint mechanism is adopted to normalize and limit the spectral energy distribution after fusion, ensuring that the energy distribution after compensation does not exhibit local over-enhancement or global imbalance. Next, by introducing a time-domain differential feedback mechanism, the current fusion result is compared with the fusion result of the previous moment to calculate the residual phase error, and this error is fed back into the update stage of the phase conjugate response function to achieve adaptive optimization. Finally, once the compensation chain is running stably, the corrected phase data is reprojected onto a unified time baseline, which dynamically strengthens the synchronization reference of the time sequence anchor group, thereby forming a two-way closed loop of phase compensation and time synchronization.

[0036] Through this process, the dual-satellite collaborative observation system has achieved a unity of phase consistency, energy conservation, and temporal stability in spatial-spectral fusion, completely eliminating spectral phase anomalies caused by time misalignment intervals, and significantly improving the authenticity of surface anomaly identification and the accuracy of spatial-spectral information fusion.

[0037] After the phase conjugate compensation chain completes real-time phase suppression, the counterfactual playback mechanism based on synchronization reference is invoked to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval. The residual false enhancement signals in the fusion result are identified and eliminated, and the differential comparison result is written back to the unified time baseline to form a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation. The specific implementation steps are as follows: After the phase conjugate compensation chain completes real-time phase suppression, to further ensure the physical consistency of the phase-compensated spatial-spectral fusion result in both the time and energy domains, a counterfactual playback mechanism based on synchronous reference is invoked to verify the authenticity and eliminate errors in the fused data. Specifically, firstly, a playback time window for the phase-compensated fused data is established using the synchronous reference time series defined by the time-series anchor group in the previous step. Within this time window, the spectral response signal at each fusion moment is mapped one-to-one with the original uncompensated signal recorded on a unified time baseline, thereby constructing spectral time series pairs before and after compensation at the same surface pixels. Then, based on this time series pair, the phase retrospective difference of the fused data is calculated, and by comparing the temporal phase gradient of the compensated spectral signal with that of historical signals, possible residual phase drift trends are identified. To improve the sensitivity of counterfactual playback, this step further introduces a time-weighted filtering method, assigning higher weights to anchor points near the edge of the time misalignment interval in the time-series anchor group, thereby amplifying the detection capability of residual misalignment signals. This process creates a continuous phase playback trajectory in the time domain, providing a high-precision dynamic reference for subsequent differential alignment.

[0038] After establishing the phase playback trajectory, the phase-compensated spatial-spectral fusion result is differentially compared with the historical trajectory data corresponding to the time misalignment interval to identify any spurious enhancement signals remaining in the fusion result. Specifically, firstly, a historical trajectory data segment corresponding to the current fusion time is selected, and time interpolation is performed on both under a unified time baseline to ensure complete alignment on the time coordinates. Next, pixel-level differential calculations are performed on the aligned spectral energy distribution. By calculating the energy gain gradient between different bands, regions abnormally higher than the statistical mean are identified. To avoid misjudgments caused by natural surface changes (such as changes in solar angle and aerosols), this step introduces spatial coherence constraints. By comparing the energy change trends of adjacent pixels, only signals exhibiting isolated enhancement in space are retained as suspected spurious enhancement targets. Subsequently, the spectral response curves of these suspected spurious enhancement targets are differentially compared with the spectral response curves at the same location in the historical trajectory, and their phase-reverse components are extracted. If the phase-reverse component is consistent with the previous compensation direction, the signal is determined to be a residual compensation amplification effect and thus marked as a spurious enhancement signal. This method can distinguish between real surface energy changes and false signals caused by algorithm residuals in the spatiotemporal joint dimension, and achieve accurate identification of false enhancement regions in the fusion results.

[0039] After identifying false enhancement signals, these signals are removed and reconstructed to restore the energy distribution of the spatial-spectral fusion result to the true level. Specifically, firstly, based on the detection results of the differential comparison stage, the set of all pixels identified as false enhancements is extracted, and an energy compensation backoff operation is performed on the spectral curve of each pixel. This operation uses phase conjugate inverse transform to subtract the conjugate value of the phase shift from the fused signal, thereby achieving energy backoff. To maintain the continuity of the spectral curve, the signal after energy backoff is smoothed in the time domain to eliminate non-physical oscillations caused by local compensation. Next, the backoff-corrected spectral curve is rematched with the synchronous reference time series provided by the time-series anchor group to ensure that the backoff signal still maintains synchronous consistency on the time baseline. To avoid excessive backoff weakening the true signal, a counterfactual probability correction strategy is introduced. Using the energy fluctuation characteristics of similar land features in historical trajectories, a reasonable range for the current backoff amplitude is calculated, and pixels exceeding the reasonable range are marked as candidate points requiring secondary verification and processed in subsequent iterative updates. Through the above multi-level backoff and rematching steps, the false enhancement signals generated by the phase compensation residuals in the fusion results are gradually weakened or eliminated, restoring the surface energy distribution to the true physical state.

[0040] After spurious enhancement signals are eliminated, the differential alignment results are written back to the unified time baseline to form a closed-loop steady-state time synchronization and energy correction mechanism throughout the entire process of dual-satellite collaborative observation. Specifically, firstly, the back-corrected spectral energy distribution is integrated with the phase residual information obtained from the differential alignment to generate a timing error correction vector. Then, this vector is written back to the unified time baseline in the form of a time index, enabling the unified time baseline to not only record time synchronization information but also dynamically reflect the energy distribution correction process over time. Before each new observation mission begins, the system automatically retrieves the time baseline data from the previous cycle and introduces the correction vector during the new drift prediction matrix generation stage, thereby achieving collaborative prediction and dynamic constraint of time and energy errors. Simultaneously, the timing anchor group updates its own time weight distribution based on the write-back results, further enhancing its synchronization reference capability for key time periods in subsequent observations.

[0041] Through this closed-loop process, the dual-satellite collaborative observation system achieves end-to-end steady-state control, from time drift prediction, phase conjugate compensation, energy differential backoff to time baseline rewriting. This steady-state time synchronization and energy correction mechanism not only effectively suppresses the residual energy amplification effect caused by time misalignment intervals, but also enables the entire spatial-spectral fusion process to achieve dynamic balance in the time, phase, and energy dimensions, thereby ensuring the physical reliability and spatiotemporal consistency of the dual-satellite collaborative observation results in surface anomaly identification, disaster monitoring, and environmental early warning.

[0042] This invention establishes a unified cross-satellite time baseline at the initial stage of dual-satellite collaborative observation and generates a drift prediction matrix with dynamic prediction capabilities by combining orbital perturbation parameters. This achieves nanosecond-level time alignment and dynamic drift correction during the spectral acquisition process of the two satellites. This method maintains continuous time consistency between the two satellites throughout the observation mission, effectively eliminating the impact of orbital perturbations and on-board clock drift on the synchronization accuracy of the spectral acquisition window. This ensures strict temporal correspondence between the acquired data from both satellites, allowing subsequent spatial-spectral fusion to be based on high-precision time consistency. This time unification mechanism significantly improves the phase consistency and energy superposition stability of multi-source data during fusion, avoiding the spread of fusion errors caused by time offsets, and providing a reliable foundation for high-fidelity detection of surface anomalies.

[0043] This invention achieves real-time suppression and closed-loop correction of spectral phase anomalies caused by temporal misalignment during spatial-spectral fusion by constructing a phase conjugate compensation chain based on a time-series anchor group and a counterfactual playback mechanism. This technical approach enables dynamic injection of phase compensation signals and differential comparison correction during the fusion process, allowing residual spurious enhancement signals to be identified and eliminated at the generation stage, thereby significantly improving the physical authenticity and energy distribution stability of the fusion results. This closed-loop time synchronization and energy correction mechanism not only effectively suppresses spurious enhancement effects caused by orbital perturbations but also enables the dual-satellite collaborative observation system to possess adaptive steady-state control capabilities, significantly improving the accuracy and reliability of surface anomaly monitoring, disaster identification, and environmental early warning.

[0044] This invention provides, for example Figure 2 The binary star collaborative multi-type anomaly spatial spectrum information fusion system shown includes a time baseline construction module, a time-slip calculation module, a phase conjugate compensation module, and a counterfactual playback correction module; The time baseline construction module establishes a unified time baseline across satellites at the beginning of the dual-satellite collaborative observation. It achieves nanosecond-level alignment of the observation time series of the two satellites through a high-precision inter-satellite communication link, and injects the orbital perturbation parameters caused by minor orbital disturbances into the unified time baseline. Based on the orbital perturbation parameters, it generates a drift prediction matrix with dynamic prediction capabilities. The time-slip solution module performs time-slip solution on the spectral acquisition windows of the two satellites under the constraint of the drift prediction matrix, extracts the potential time misalignment intervals caused by small orbital perturbations, and generates a time-series anchor point group based on the time misalignment intervals; The phase conjugate compensation module constructs a phase conjugate compensation chain based on the time-series anchor group, performs inversion modeling of spectral phase anomalies caused by time misalignment intervals, and dynamically injects phase compensation signals into the spatial-spectral fusion channel, so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process. The counterfactual playback correction module, after the phase conjugate compensation chain completes real-time phase suppression, calls the counterfactual playback mechanism based on synchronization reference to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval, identify and remove residual false enhancement signals in the fusion result, and write the differential comparison result back to the unified time baseline, forming a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation.

[0045] The binary star-coordinated multi-type anomaly spatial spectrum information fusion method provided in this embodiment of the invention is implemented through the aforementioned binary star-coordinated multi-type anomaly spatial spectrum information fusion system. For details of the specific methods and processes of the binary star-coordinated multi-type anomaly spatial spectrum information fusion system, please refer to the embodiments of the binary star-coordinated multi-type anomaly spatial spectrum information fusion method, which will not be repeated here.

[0046] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for fusing multi-type anomaly spatial spectrum information through binary star collaboration, characterized in that, Includes the following steps: In the initial stage of dual-satellite collaborative observation, a unified time baseline is established across satellites. Nanosecond-level alignment of the observation time series of the two satellites is achieved through a high-precision inter-satellite communication link. Orbital perturbation parameters caused by minor orbital disturbances are injected into the unified time baseline. A drift prediction matrix with dynamic prediction capabilities is generated based on the orbital perturbation parameters. Under the constraint of the drift prediction matrix, time-slip calculation is performed on the spectral acquisition windows of the two satellites to extract the potential time misalignment intervals caused by small orbital perturbations, and a time-series anchor point group is generated based on the time misalignment intervals. A phase conjugate compensation chain is constructed based on the time-series anchor group to invert and model the spectral phase anomaly caused by the time misalignment interval. The phase compensation signal is dynamically injected into the spatial-spectral fusion channel so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process. After the phase conjugate compensation chain completes real-time phase suppression, the counterfactual playback mechanism based on synchronization reference is invoked to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval. The residual false enhancement signals in the fusion result are identified and eliminated, and the differential comparison result is written back to the unified time baseline, forming a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation.

2. The method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to claim 1, characterized in that, The steps to establish a unified time baseline across satellites include: The ground control center issues a time synchronization command, establishes a two-way time calibration channel through a high-precision inter-satellite communication link between satellites, and uses the difference in round-trip ranging delay for correction. The orbital perturbation parameters caused by minor orbital disturbances are injected into a unified time baseline to give it dynamic response characteristics; A drift prediction matrix with dynamic prediction capability is generated based on the orbital perturbation parameters. Weighted fitting and recursive prediction are performed on each perturbation component to obtain the estimated value of time drift. During the spectral acquisition phase, the spectral acquisition trigger time is shifted and corrected based on the prediction results in the drift prediction matrix, and closed-loop time correction is achieved through inter-satellite link round-trip delay feedback.

3. The method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to claim 2, characterized in that, In the process of generating the drift prediction matrix, the weighted fitting of the orbital perturbation parameters adopts a combination of Kalman filtering and polynomial extrapolation to achieve continuous prediction of the time drift. During satellite operation, the internal weight coefficients are automatically adjusted based on the real-time feedback of the inter-satellite link round-trip delay.

4. The method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to claim 1, characterized in that, The steps for performing time-slip solution under the constraints of the drift prediction matrix include: The drift prediction matrix is ​​jointly matched with the on-orbit attitude information, imaging time stamps and ground reference time series of the two satellites to dynamically correct the observation time and calculate the initial estimate of the time offset. A weighted least squares sliding fitting algorithm is introduced to approximate the time series piecewise and impose continuous constraints, thereby obtaining the time sliding solution results; Using the drift sensitivity parameter of the drift prediction matrix as a constraint, the potential time misalignment interval caused by orbital disturbance is extracted and a time misalignment parameter set is formed; A time-series anchor group is generated based on a time-misaligned parameter set. The anchor time is fine-tuned and dynamically corrected through coupling verification with the drift prediction matrix and the time-slip solution, ensuring the temporal continuity and spatial consistency of the time-series anchor group.

5. The binary star cooperative multi-type anomaly spatial spectrum information fusion method according to claim 4, characterized in that, In the step of generating the time-series anchor group, the time coordinates of each anchor point are compared with the predicted drift values ​​in the drift prediction matrix. When the deviation exceeds the allowable threshold, the anchor point time is fine-tuned based on the time-slip solution results, and real-time correction is performed through the dynamic prediction model of the drift prediction matrix.

6. The method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to claim 1, characterized in that, The steps for constructing a phase conjugate compensation chain based on a time-series anchor group include: The time-series anchor point group is time-domain aligned with the spectral acquisition data of the two satellites in the same surface area, and the spectral phase shift function is calculated based on the time weight and spatial coordinates of each anchor point. The phase shift component is extracted based on the spectral phase shift function and spectral decomposition is performed. Inversion equations for the low-frequency trend term, high-frequency disturbance term, and transient phase change term are established respectively, generating a continuous phase conjugate response function. By embedding the phase conjugate response function into the spatial-spectral fusion channel, dynamic injection of phase compensation signal is achieved, and real-time spectral phase correction is realized through adaptive weight allocation of land cover categories. After compensation, real-time monitoring is performed using a sliding observation window of the time-series anchor group, and the weights of the phase conjugate response function are corrected by the drift prediction matrix.

7. The binary star cooperative multi-type anomaly spatial spectrum information fusion method according to claim 6, characterized in that, In the process of embedding the phase conjugate response function into the spatial-spectral fusion channel and realizing the dynamic injection of the phase compensation signal, a phase energy constraint mechanism is introduced to normalize and limit the spectral energy distribution after fusion, and a time-domain differential feedback mechanism is combined to adaptively optimize the weight of the phase conjugate response function to prevent excessive energy enhancement or phase drift accumulation during the compensation process.

8. The method for fusing multi-type anomaly spatial spectrum information in binary star collaboration according to claim 1, characterized in that, The steps to invoke a counterfactual replay mechanism based on synchronization references include: A playback time window for the fused data after phase compensation is established using the synchronous reference time series defined by the time anchor group, and the phase backtracking difference of the spectral signals before and after compensation is calculated to form the phase playback trajectory. The spatial spectrum fusion result after phase compensation is compared with the historical trajectory data corresponding to the time misalignment interval to identify and mark the residual false enhancement signals. Based on the differential alignment results, an energy backoff operation is performed on the spurious enhanced signal, and the spectral signal is rematched and stably restored through a counterfactual probability correction strategy; The backtracked spectral energy distribution and phase residual information are integrated to generate a timing error correction vector, which is then written back to the unified time baseline.

9. The method for fusing multi-type anomalous spatial spectrum information in binary star collaboration according to claim 8, characterized in that, In the energy backoff operation, the phase conjugate inverse transform is used to subtract the conjugate value of the phase shift from the fused signal, and time domain smoothing is used after backoff to eliminate local non-physical oscillations. At the same time, a counterfactual probability correction threshold is set based on the energy fluctuation characteristics of similar land features in historical trajectories to limit the energy backoff amplitude.

10. A binary star-coordinated multi-type anomaly spatial spectrum information fusion system, used to implement the binary star-coordinated multi-type anomaly spatial spectrum information fusion method according to any one of claims 1-9, characterized in that, It includes a time baseline construction module, a time slip calculation module, a phase conjugate compensation module, and a counterfactual playback correction module; The time baseline construction module establishes a unified time baseline across satellites at the beginning of the dual-satellite collaborative observation. It achieves nanosecond-level alignment of the observation time series of the two satellites through a high-precision inter-satellite communication link, and injects the orbital perturbation parameters caused by minor orbital disturbances into the unified time baseline. Based on the orbital perturbation parameters, it generates a drift prediction matrix with dynamic prediction capabilities. The time-slip solution module performs time-slip solution on the spectral acquisition windows of the two satellites under the constraint of the drift prediction matrix, extracts the potential time misalignment intervals caused by small orbital perturbations, and generates a time-series anchor point group based on the time misalignment intervals; The phase conjugate compensation module constructs a phase conjugate compensation chain based on the time-series anchor group, performs inversion modeling of spectral phase anomalies caused by time misalignment intervals, and dynamically injects phase compensation signals into the spatial-spectral fusion channel, so that the synchronization reference corresponding to the time-series anchor group is strengthened in real time during the fusion process. The counterfactual playback correction module, after the phase conjugate compensation chain completes real-time phase suppression, calls the counterfactual playback mechanism based on synchronization reference to perform differential comparison between the phase-compensated spatial spectrum fusion result and the historical trajectory data corresponding to the time misalignment interval, identify and remove residual false enhancement signals in the fusion result, and write the differential comparison result back to the unified time baseline, forming a closed-loop steady-state time synchronization and energy correction mechanism that runs through the entire process of dual-star collaborative observation.

Images