Data assimilation method, device and equipment

Abstract

The invention relates to the technical field of detection data assimilation, in particular to a data assimilation method, device and equipment, and the method comprises the steps: carrying out the multi-beam scanning data collection according to a detection system, so as to obtain a detection state data set; performing dynamic target detection and tracking according to the detection state data set to obtain detection target movement track data; spatial distribution feature analysis is carried out on the detection target motion trail data to obtain a target group spatial distribution feature data set, and the target group spatial distribution feature data set comprises spatial density distribution parameters and velocity field gradient features of a target group. According to the method, through multi-beam scanning data acquisition and dynamic target detection tracking, the motion trail data of the target can be more accurately acquired, and through analysis of the spatial distribution characteristics of the target group, including spatial density distribution parameters and velocity field gradient characteristics, the distribution and dynamic behaviors of the target can be more clearly understood.

Description

Technical Field

[0001] The present invention relates to the technical field of detection data assimilation, and in particular to a data assimilation method, device and equipment. Background Art

[0002] Most traditional methods rely on classical physics models, mainly including detection wave propagation models and target motion models. These models assume that the target motion is linear or quasi-linear and that wave propagation follows certain idealized rules. Among them, the detection wave propagation model is used to assume that when the detection wave propagates, environmental factors (such as atmospheric refraction, ionosphere, etc.) have little effect on the signal, and ignores the correction of complex meteorological conditions and ionospheric disturbances. The target motion model is used to assume that the target moves according to a known pattern (such as a straight line or uniform motion), and usually uses simple velocity and acceleration formulas to describe the target trajectory.

[0003] Based on the above content, it is clear that traditional methods are often unable to accurately capture the motion trajectory and distribution characteristics of the target, especially in complex environments (such as atmospheric refraction, ionospheric interference, etc.). Due to the lack of in-depth analysis of the spatial distribution characteristics of the target group, traditional methods may cause large data deviations due to observation errors, thereby affecting the accuracy of the prediction; and when processing detection status data, traditional methods usually lack corrections for factors such as atmospheric refractive index, meteorological changes, and ionospheric disturbances, which may lead to errors in propagation paths and echo signals; and in traditional methods, the assimilation processing of detection data is relatively simplified, and the application of error compensation and correction factors is often ignored, which makes it impossible to effectively refine the spatial distribution characteristic data of the target group, resulting in high uncertainty in the model results. Moreover, traditional methods have weak initial state optimization capabilities for numerical weather forecast systems and are difficult to cope with rapid changes in weather patterns, resulting in low prediction accuracy. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a data assimilation method, device and equipment.

[0005] The technical solution adopted to solve the above technical problems is: a data assimilation method, including: Performing multi-beam scanning data acquisition according to the detection system to obtain a detection state data set; performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data; Performing spatial distribution feature analysis on the detected target motion trajectory data to obtain a target group spatial distribution feature data set, wherein the target group spatial distribution feature data set includes spatial density distribution parameters and velocity field gradient characteristics of the target group; Acquiring real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; performing electromagnetic propagation effect modeling on the detection status dataset based on the real-time meteorological data to obtain a propagation path correction factor; Performing observation error compensation on the target group spatial distribution characteristic data set according to the propagation path correction factor to obtain error-corrected target group spatial distribution characteristic data; An assimilation analysis model is constructed for the error-corrected target group spatial distribution characteristic data using an ensemble Kalman filter algorithm to obtain a data assimilation model; and the data assimilation model is deployed in a numerical weather prediction system.

[0006] Preferably, performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data includes: performing a time domain convolution operation on the transmission signal and the detection state data set to obtain high-resolution range image data; performing frequency domain filtering on the high-resolution range image data according to a Doppler filter bank to separate the moving target echo from the stationary clutter; Setting a constant false alarm rate detection threshold to perform target point detection on the moving target echo to obtain an initial target point set; Establishing a detection unit spatial index in a polar coordinate system according to the detection beam pointing parameter, and mapping the initial target point trace set to a grid coordinate system of a detection scanning sector according to the detection unit spatial index to obtain a time-space aligned initial target point trace set; A time series alignment model is constructed according to the detection scanning cycle parameters, and an initial target point trace set obtained by different beam pointings is projected into a unified Cartesian coordinate system through coordinate transformation according to the time series alignment model to obtain a time-space aligned point trace set; Initialize the trajectory of the newly appeared target to obtain a candidate trajectory set, wherein the candidate trajectory set includes an unconfirmed trajectory and a first confirmed trajectory.

[0007] Preferably, performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data further includes: Performing interactive multi-model filtering on the first confirmed trajectory to obtain a state prediction value of the target, wherein the interactive multi-model filtering adopts a hybrid state transition model of a uniform speed model, a uniform acceleration model, and a coordinated turning model; State prediction and updating are performed based on Kalman filtering, and the spatial coordinates, radial velocity, and signal-to-noise ratio information of the spatiotemporal alignment point trace set are integrated to obtain an optimized state prediction value of the target; The process noise covariance matrix and the measurement noise covariance matrix are dynamically modified according to the target echo intensity to obtain the adaptive filtering parameter set; the trajectory confirmation threshold and deletion time limit are set to obtain the trajectory status label set.

[0008] Preferably, performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data further includes: Establishing an association matrix based on the state prediction value of the target appearance and the spatial proximity and Doppler frequency consistency of the spatiotemporal aligned point trace set, and determining an association probability between the spatiotemporal aligned point trace set and the first confirmed trajectory based on the association matrix; Performing a weighted summation of the length of the first confirmed trajectory and the association probability to obtain a trajectory confidence score; screening the first confirmed trajectory according to the trajectory confidence score to obtain a second confirmed trajectory; The spatial coordinate sequence of the second confirmed trajectory is converted into an Earth-centered fixed coordinate system to obtain detection target motion trajectory data.

[0009] Preferably, performing spatial distribution feature analysis on the detected target motion trajectory data to obtain a target group spatial distribution feature data set includes: According to the three-dimensional spatial coordinate sequence of the detection target motion trajectory data, a target point trace is spatially distributed modeled by a kernel density estimation algorithm to obtain a density distribution surface; Identifying local density extreme value areas in the density distribution surface according to a density peak clustering algorithm to obtain distribution area parameters of the target group, wherein the distribution area parameters include peak density, average density, and contour line enclosed area; Performing regional classification on the density distribution surface according to the distribution area parameters of the target group to obtain high-density areas and low-density areas, and performing noise filtering based on neighborhood connectivity on the low-density areas to obtain a set of spatial density distribution parameters; constructing a three-dimensional velocity vector field according to the velocity vector sequence of the detection target motion trajectory data, and determining a velocity gradient tensor of the three-dimensional velocity vector field according to a five-point central difference method; A convolution operation is performed on the velocity gradient tensor according to a Sobel operator to obtain velocity field gradient features, wherein the velocity field gradient features include maximum gradient amplitude, gradient direction variance, and velocity field curvature.

[0010] Preferably, performing electromagnetic propagation effect modeling on the detection status data set according to the real-time meteorological data to obtain a propagation path correction factor includes: Constructing a three-dimensional atmospheric refractive index field using a Kriging interpolation algorithm based on the atmospheric refractive index profile data, and combining the detection site location and detection range to obtain a spatially continuously distributed refractive index model; Numerical integration is performed according to the fourth-order Runge-Kutta method to obtain a region where the refractive index gradient changes dramatically; and the actual propagation path of the electromagnetic wave in the region where the refractive index gradient changes dramatically is determined according to a three-dimensional ray tracing equation; determining a geometric correction factor according to a deviation between the actual propagation path of the electromagnetic wave and the apparent path, and determining a path delay compensation amount according to the geometric correction factor; constructing an ionospheric equivalent phase screen model according to the ionospheric disturbance parameters, wherein the ionospheric equivalent phase screen model includes a total electron content and a scintillation index, and describes the spatial distribution of the phase disturbance according to a piecewise linear method; determining a phase delay and a phase scintillation caused by the ionosphere according to the spatial distribution of the phase disturbance and the ionosphere equivalent phase screen model, and determining a phase compensation factor according to the phase delay and the phase scintillation; The path delay compensation amount and the phase compensation factor are vectorially superimposed to obtain a propagation path correction factor.

[0011] Preferably, an ensemble Kalman filter algorithm is used to construct an assimilation analysis model for the error-corrected target group spatial distribution characteristic data to obtain a data assimilation model, including: The background field data of the numerical weather prediction model is projected onto the spatiotemporal coordinate system of the detection observation space according to a bilinear interpolation algorithm to obtain a spatiotemporally aligned background field data set, wherein the background field data includes wind field, temperature field, and pressure field; The error-corrected target group spatial distribution characteristic data is transformed into variables to obtain an equivalent momentum flux field; the perturbation is generated through the background field error covariance matrix according to singular value decomposition to obtain an initial perturbation set; The disturbance amplitude of the initial disturbance set is matched with the uncertainty magnitude of the background field to obtain the initial state set of the numerical weather prediction model; random error disturbance is applied to the detection observation field to obtain the observation set.

[0012] Preferably, the assimilation analysis model is constructed by using an ensemble Kalman filter algorithm to perform assimilation analysis on the error-corrected target group spatial distribution characteristic data to obtain a data assimilation model, further comprising: Performing a numerical weather prediction model integration on each member of the initial state set to predict the atmospheric state at a future time, thereby obtaining a predicted state set; mapping the predicted state set to a detection observation space, determining an equivalent predicted observation quantity, thereby obtaining a predicted observation set; Determining a covariance matrix of the predicted state set and a covariance matrix of the predicted observation set; determining a correction weight of the observation data on the state of the numerical weather prediction model based on covariance analysis to obtain a Kalman gain matrix; The predicted state set is dynamically adjusted according to the Kalman gain matrix, and the constraint information of the detection observation data is combined to obtain the analysis state set; the error covariance matrix is ​​updated according to the statistical characteristics of the analysis state set to obtain the data assimilation model.

[0013] The technical solution adopted to solve the above technical problems is: a data assimilation device, which is applicable to the above-mentioned data assimilation method, comprising: A data acquisition module, the data acquisition module is used to perform multi-beam scanning data acquisition according to the detection system to obtain a detection status data set; perform dynamic target detection and tracking according to the detection status data set to obtain detection target motion trajectory data; A feature analysis module, wherein the feature analysis module is used to perform spatial distribution feature analysis on the detection target motion trajectory data to obtain a target group spatial distribution feature data set, wherein the target group spatial distribution feature data set includes spatial density distribution parameters and velocity field gradient characteristics of the target group; a propagation correction module, the propagation correction module being configured to acquire real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; and perform electromagnetic propagation effect modeling on the detection status dataset based on the real-time meteorological data to obtain a propagation path correction factor; an error compensation module, the error compensation module being used to perform observation error compensation on the target group spatial distribution feature data set according to the propagation path correction factor to obtain error-corrected target group spatial distribution feature data; A data assimilation module is used to construct an assimilation analysis model for the error-corrected target group spatial distribution characteristic data through an ensemble Kalman filter algorithm to obtain a data assimilation model; and deploy the data assimilation model to a numerical weather prediction system.

[0014] The technical solution adopted to solve the above technical problems is: a data assimilation device, including: a processor and a memory; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, so that the processor performs the data assimilation method.

[0015] The beneficial effects of the present invention are as follows: (1) The present invention can more accurately obtain the target's motion trajectory data through multi-beam scanning data acquisition and dynamic target detection and tracking, and through the analysis of the spatial distribution characteristics of the target group, including spatial density distribution parameters and velocity field gradient characteristics, it helps to more clearly understand the distribution and dynamic behavior of the target, thereby reducing data deviations caused by observation errors, and by obtaining real-time meteorological data, especially atmospheric refractive index profile data and ionospheric disturbance parameters, and correcting the propagation path of the detection status data through electromagnetic propagation effect modeling, it can effectively reduce the errors caused by environmental factors (such as atmospheric refraction, ionospheric disturbance) and ensure the accuracy of the detection data; (2) The present invention can effectively reduce the errors caused by environmental factors (such as atmospheric refraction, ionospheric disturbance) through the use of multi-beam scanning data acquisition and dynamic target detection and tracking, thereby ensuring the accuracy of the detection data. By using the propagation path correction factor to compensate for the error of the target group spatial distribution characteristic data, more accurate target group spatial distribution characteristic data can be obtained, which is crucial for the assimilation of detection data and helps to improve the credibility and accuracy of the model. By assimilation analysis of the error-corrected target group spatial distribution characteristic data through the ensemble Kalman filter algorithm, the initial state in the numerical weather prediction system can be optimized, the uncertainty of the model can be reduced, and thus the accuracy of the prediction can be improved, especially when the weather pattern changes rapidly. By effectively combining the data assimilation model with the numerical weather prediction system, the detection data can be used to optimize the weather forecast in real time, and the ability to predict meteorological changes can be improved, especially in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of the steps of the overall method in one embodiment of the present invention; Figure 2 A schematic diagram of the overall device architecture of an embodiment of the present invention; Figure 3 This is a schematic diagram of the device architecture of an overall device in an embodiment of the present invention.

[0017] Figure numerals: 1. Data acquisition module; 2. Feature analysis module; 3. Propagation correction module; 4. Error compensation module; 5. Data assimilation module; 6. Processor; 7. Memory. DETAILED DESCRIPTION

[0018] Example 1, as Figure 1 As shown, the present invention proposes a data assimilation method, which is characterized by comprising: S1. Perform multi-beam scanning data acquisition based on the detection system to obtain a detection status data set; perform dynamic target detection and tracking based on the detection status data set to obtain detection target motion trajectory data; S2. Analyzing the spatial distribution characteristics of the detected target motion trajectory data to obtain a target group spatial distribution characteristic dataset, wherein the target group spatial distribution characteristic dataset includes spatial density distribution parameters and velocity field gradient characteristics of the target group; S3. Acquire real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; perform electromagnetic propagation effect modeling on the detection status data set based on the real-time meteorological data to obtain a propagation path correction factor; S4. Compensating the target group spatial distribution characteristic data set for observation errors according to the propagation path correction factor to obtain error-corrected target group spatial distribution characteristic data; S5. Use the ensemble Kalman filter algorithm to construct an assimilation analysis model for the error-corrected spatial distribution characteristic data of the target group to obtain a data assimilation model; and deploy the data assimilation model to the numerical weather prediction system.

[0019] In the present invention, observation error compensation is performed on the target group spatial distribution characteristic data set according to the propagation path correction factor to obtain the target group spatial distribution characteristic data after error correction, specifically as follows: spatiotemporal interpolation processing is performed on the propagation path correction factor, and the spatial resolution of the correction factor field is improved to be consistent with the target group distribution data set by using the Kriging algorithm. The time dimension is synchronized at the second level through cubic spline interpolation to ensure that the correction factor is strictly in time and space with the target group trajectory points; a conversion matrix from the detection coordinate system to the Earth-centered fixed coordinate system (ECEF) is constructed, and the spatial reference differences between different data sources are eliminated through a seven-parameter coordinate transformation model to achieve spatial alignment of the correction factor with the target group trajectory points; the atmospheric refraction delay term and the ionospheric projection deviation term in the analytical geometry correction factor are used to establish a position correction model: for the path bending effect caused by atmospheric refraction, the target position is corrected by inverse ray tracing operation, and the distance deviation is compensated along the detection line of sight direction; for the position offset caused by the inhomogeneity of the ionospheric electron density, the ionospheric puncture point (IPP) projection method is used for reverse correction to eliminate the azimuth deviation caused by the ionospheric refraction; RA is performed on the compensated position data The NSAC algorithm is used for filtering to eliminate position jumps caused by abnormal correction factors, while retaining a minimum of 95% of inliers to ensure continuity of the spatial distribution. The ionospheric delay and scintillation disturbance terms in the phase compensation factor are extracted to construct a velocity correction model. For velocity deviations caused by ionospheric dispersion, the influence of first-order terms is eliminated through carrier frequency normalization. For velocity fluctuations caused by phase scintillation, an adaptive threshold method is used to identify and filter out high-frequency disturbances, preserving valid velocity information. A Kalman smoother is used to filter the velocity corrections in the time domain, setting a process noise covariance matrix to suppress high-frequency noise while preserving the dynamic characteristics of small and medium-scale weather systems. Based on the compensated position data, the spatial density distribution is recalculated using the adaptive bandwidth kernel density estimation (ABKDE). The bandwidth parameter is dynamically adjusted using the Silverman rule to ensure a resolution better than 500 meters in high-density areas while smoothing noise in low-density areas. Cluster analysis of the density field is performed using the improved DBSCAN algorithm, with thresholds for the neighborhood radius and number of core points set to identify the core distribution area of ​​the corrected target cluster and filter out spurious density peaks introduced by error compensation.

[0020] In the second embodiment, a data assimilation method proposed by the present invention, compared with the first embodiment, this embodiment further includes: performing dynamic target detection and tracking based on the detection state data set to obtain detection target motion trajectory data, including: A1. Perform a time-domain convolution operation on the transmitted signal and the detection status data set to obtain high-resolution range image data. Perform frequency-domain filtering on the high-resolution range image data using a Doppler filter bank to separate the moving target echo from the stationary clutter. A2. Setting a constant false alarm rate detection threshold to perform target point detection on the moving target echo to obtain an initial target point set; A3. Establishing a detection unit spatial index in a polar coordinate system based on the detection beam pointing parameters. Mapping the initial target point trace set to the grid coordinate system of the detection scanning sector based on the detection unit spatial index to obtain a spatiotemporally aligned initial target point trace set. A4. Construct a time series alignment model based on the detection scan cycle parameters. Project the initial target point traces obtained by different beam pointing directions into a unified Cartesian coordinate system through coordinate transformation based on the time series alignment model to obtain a spatiotemporally aligned point trace set. A5. Initialize the trajectory of the newly appeared target to obtain a candidate trajectory set, where the candidate trajectory set includes an unconfirmed trajectory and a first confirmed trajectory.

[0021] In this embodiment, a convolution operation is performed on the transmitted signal and the received detection state data in the time domain. Time domain convolution is a signal processing method used to extract the features of the detection state data, and ultimately obtain high-resolution range image data, which contains the distance information of the target in the detection scanning area; the high-resolution range image data is subjected to frequency domain filtering, the purpose of which is to separate the echo of the moving target from the stationary clutter through the Doppler effect; the Doppler filter group distinguishes the echo signals of the moving target and the stationary object according to different frequency changes, ensuring that only the real moving target is focused on; a constant false alarm rate (CFAR) detection threshold is set to detect whether there is a target in the detection state; the CFAR method can automatically adjust the threshold value according to the background clutter signal to avoid false detection of environmental noise; when the target echo signal is received, the target position can be confirmed, thereby obtaining an initial target point track set; the index of the detection unit space is established in the polar coordinate system through the beam pointing parameters of the phased array detection; because the area scanned by the phased array detection is The domain is usually fan-shaped, so the target points are spatially divided and indexed in the polar coordinate system for subsequent processing; the initial target point set is mapped to the grid coordinate system of the detection scan to achieve spatiotemporal alignment; this step ensures that the target point set corresponds to the sector of the detection scan, which is convenient for further analysis; a time series alignment model is constructed based on the detection scan cycle parameters; this means that as the detection scan cycle changes, the time information of the target point traces needs to be aligned with the points under different beam directions; through coordinate transformation, these target point traces are projected into a unified Cartesian coordinate system to obtain a spatiotemporal aligned point set; the track of the newly appeared target is initialized; this means that if a new target is detected, the system needs to make a preliminary estimate of its motion trajectory and initialize the trajectory; the candidate trajectory set includes two types of trajectories: one is an unconfirmed trajectory (which may be a false detection or a temporary target); the other is the first confirmed trajectory (a confirmed target); the candidate trajectory set helps to subsequently determine the true trajectory of the target.

[0022] In an optional embodiment, dynamic target detection and tracking are performed based on the detection state data set to obtain detection target motion trajectory data, further comprising: A6. Performing interactive multi-model filtering on the first confirmed trajectory to obtain a state prediction value of the target, wherein the interactive multi-model filtering uses a hybrid state transition model of a uniform velocity model, a uniform acceleration model, and a coordinated turning model; A7. Perform state prediction and update based on Kalman filtering, integrating the spatial coordinates, radial velocity, and signal-to-noise ratio information of the spatiotemporal alignment point set to obtain the optimized state prediction value of the target; A8. Dynamically modify the process noise covariance matrix and the measurement noise covariance matrix according to the target echo intensity to obtain an adaptive filtering parameter set; set the trajectory confirmation threshold and deletion time limit to obtain a trajectory status label set.

[0023] It should be noted that interactive multi-model filtering is a filtering method based on multiple different motion models, which can handle targets with different dynamic behaviors; here, IMM uses three different models: uniform speed model: assuming that the target moves at a constant speed; uniform acceleration model: assuming that the target's movement is uniformly accelerated, that is, the target's speed changes continuously during the movement; collaborative turning model: used to describe the situation where the target turns during movement, suitable for the target's turning trajectory; these models are combined through a hybrid state transfer model, and the filter will select different models according to the target's movement to obtain more accurate state prediction; state prediction and update based on Kalman filter: Kalman filter is a recursive filtering algorithm based on a linear system, used to estimate the state of the system; in this application, Kalman filter combines the target's spatiotemporal aligned point set (including spatial coordinates, radial velocity and signal-to-noise ratio information) and performs state prediction and update based on these data; the Kalman filter is used to predict the target's state, that is, based on the target's current state and motion model, its future position and velocity are predicted; by comparing with the actual measurement value, the predicted value is updated to reduce the error The acoustic covariance matrix is ​​used to describe the magnitude and relationship of process noise and measurement noise in the system. In this method, the noise covariance matrix is ​​dynamically adjusted based on the target echo intensity to ensure that the filter can adapt to the target detection environment under different signal-to-noise ratio conditions. The process noise covariance matrix describes the uncertainty of the system itself, such as the dynamic model of the target, and describes the errors that may occur during the measurement process, such as detection noise or external interference. Based on the dynamically adjusted noise covariance matrix, an adaptive filter parameter set can be obtained. This means that the filter can self-adjust its parameters according to the actual echo intensity and environmental changes, thereby improving the accuracy of target state prediction. It is used to determine whether a target trajectory is valid. When the target state remains stable for a period of time and exceeds a certain threshold, the trajectory is considered valid and reliable. If the target does not update within a certain period of time or its state no longer meets expectations, the trajectory will be deleted according to the deletion time limit rule to avoid interference from false detection or invalid trajectories. Finally, the system generates a trajectory state label set to mark the trajectory state of each target. These labels usually include information such as whether the target is confirmed and whether the trajectory is valid, which is used for subsequent processing and decision-making.

[0024] In an optional embodiment, dynamic target detection and tracking are performed based on the detection state data set to obtain detection target motion trajectory data, further comprising: A8. Establishing a correlation matrix based on the state prediction value of the target and the spatial proximity and Doppler frequency consistency of the spatiotemporally aligned point trace set, and determining the probability of association between the spatiotemporally aligned point trace set and the first confirmed track based on the correlation matrix; A9. Perform a weighted summation of the length and the associated probability of the first confirmed trajectory to obtain a trajectory confidence score; filter the first confirmed trajectory according to the trajectory confidence score to obtain a second confirmed trajectory; A10. Convert the spatial coordinate sequence of the second confirmed trajectory into a geocentric fixed coordinate system to obtain detection target motion trajectory data.

[0025] It should be noted that the target state prediction value predicts the target state (position, speed, etc.) through the target's dynamic model (such as Kalman filtering); the spatiotemporal aligned point track set usually refers to a set of target point track data obtained through detection or other sensors, which can provide the target's specific position information in time and space after spatiotemporal alignment processing; the two parameters of spatial proximity and Doppler frequency consistency are used to evaluate the relationship between each point track and the target; spatial proximity refers to the spatial proximity between the target predicted position and the point track, and Doppler frequency consistency refers to whether the target's frequency offset matches the predicted target frequency; this information is used to establish an association matrix, and each element in the matrix represents the degree of association between the point track and the target trajectory; based on the association matrix established above, the association probability between the point track and the confirmed trajectory (i.e., the first confirmed trajectory) can be calculated; this probability represents the association between a spatiotemporal aligned point track set and the confirmed The correlation between trajectories; the length of a confirmed trajectory refers to the time span or distance of the target trajectory; a longer trajectory usually means that the target moves stably within a certain period of time and has a higher confidence level; the confidence score of the target trajectory can be obtained by weighted summation of the association probability with the point trace; this score measures the reliability of the target trajectory, and the higher the score, the more likely the trajectory is correct; based on the calculated confidence score, a new confirmed trajectory (i.e., the second confirmed trajectory) is screened out; a high-confidence trajectory means that the target's motion trajectory is more stable and can continue to be tracked; a low-confidence trajectory may be discarded or require further verification; finally, the spatial coordinate sequence of the second confirmed trajectory will be converted to a geocentric fixed coordinate system, that is, the trajectory is converted from a ground coordinate system to a coordinate system with the center of the earth as the origin; this conversion allows the trajectory data to more accurately represent the position and movement of the target on the surface of the earth; In an optional embodiment, performing spatial distribution feature analysis on the detected target motion trajectory data to obtain a target group spatial distribution feature dataset includes: B1. Based on the three-dimensional spatial coordinate sequence of the detected target motion trajectory data, the target point traces are spatially distributed modeled using a kernel density estimation algorithm to obtain a density distribution surface; B2. Identify local density extreme value areas in the density distribution surface using a density peak clustering algorithm to obtain distribution area parameters of the target group, where the distribution area parameters include peak density, average density, and contour area. B3. Classify the density distribution surface according to the distribution area parameters of the target group to obtain high-density areas and low-density areas. Perform noise filtering based on neighborhood connectivity on the low-density areas to obtain a set of spatial density distribution parameters. B4. Construct a three-dimensional velocity vector field based on the velocity vector sequence of the detected target motion trajectory data, and determine the velocity gradient tensor of the three-dimensional velocity vector field using the five-point central difference method; B5. Perform a convolution operation on the velocity gradient tensor according to the Sobel operator to obtain velocity field gradient features, where the velocity field gradient features include maximum gradient amplitude, gradient direction variance, and velocity field curvature.

[0026] It should be noted that kernel density estimation is used to analyze the three-dimensional trajectory data of the detected target, and a density distribution surface is established by calculating the spatial density distribution of the points. This surface shows the density of the target in different areas in space and reflects the distribution characteristics of the target points. Density peak clustering is a clustering algorithm that determines the cluster center by finding the local extreme points of density in the data. Specifically, the algorithm identifies the local density extreme areas in the density distribution surface, thereby dividing the distribution area of ​​the target group. The distribution area parameters include: peak density refers to the highest density value in the area; average density refers to the average density of the points in the area; the contour area refers to the size of the area around the peak density, usually expressed as an equal density area. According to the distribution area parameters of the target group, the density distribution surface is classified into high-density areas and low-density areas. For low-density areas, noise filtering is performed through the neighborhood connectivity method to remove areas with too low density and possible noise. This step helps to reduce erroneous target points and thus obtain a more accurate target trajectory; the velocity vector field represents the direction and velocity of each point, reflecting the motion information of the target; by detecting the target's trajectory data, a three-dimensional velocity vector field is constructed; the velocity gradient tensor is an important quantity that describes the change in the velocity field; the gradient of the velocity vector field is determined by the five-point central difference method (that is, by calculating the difference between adjacent points); this can capture the spatial characteristics of the target velocity change; the Sobel operator is an edge detection algorithm commonly used in image processing, which can extract gradient information from the image; in this process, the Sobel operator is used to convolve the velocity gradient tensor to obtain the gradient characteristics of the velocity field; the velocity field gradient characteristics include: maximum gradient amplitude: the maximum change amplitude of the gradient in the velocity field; gradient direction variance: the degree of change in the gradient direction, used to describe the irregularity of the velocity field; velocity field curvature: the curvature of the velocity field change, reflecting the degree of curvature of the velocity field.

[0027] In an optional embodiment, electromagnetic propagation effect modeling is performed on the detection status dataset based on real-time meteorological data to obtain a propagation path correction factor, including: C1. Construct a three-dimensional atmospheric refractivity field using the Kriging interpolation algorithm based on atmospheric refractivity profile data. Combined with the detection station location and detection range, a spatially continuous refractivity model is obtained. C2. Perform numerical integration based on the fourth-order Runge-Kutta method to obtain the region where the refractive index gradient changes dramatically; determine the actual propagation path of the electromagnetic wave in the region where the refractive index gradient changes dramatically based on the three-dimensional ray tracing equation; C3. Determine a geometric correction factor based on the deviation between the actual propagation path of the electromagnetic wave and the apparent path, and determine a path delay compensation amount based on the geometric correction factor; C4. Based on the ionospheric disturbance parameters, an ionospheric equivalent phase screen model is constructed. The ionospheric equivalent phase screen model includes the total electron content and the scintillation index, and the spatial distribution of the phase disturbance is described using a piecewise linear method. C5. Determine the phase delay and phase scintillation caused by the ionosphere based on the spatial distribution of phase disturbances and the ionosphere equivalent phase screen model, and determine the phase compensation factor based on the phase delay and phase scintillation; C6. Perform vector superposition of the path delay compensation amount and the phase compensation factor to obtain a propagation path correction factor.

[0028] It should be noted that the Kriging interpolation algorithm is a statistical method used to estimate the value of spatial data based on the spatial distribution of known data points. For atmospheric refractive index profile data, the Kriging interpolation algorithm can construct a three-dimensional refractive index field based on the irregularly distributed refractive index data in the atmosphere. The refractive index field reflects the refractive index distribution in different regions of the atmosphere and is used to describe the propagation characteristics of electromagnetic waves in the atmosphere. Combined with the location and detection range of the detection station, the three-dimensional refractive index field provides a continuous distribution of atmospheric refractive index in space, which helps to predict the propagation behavior of electromagnetic waves in different atmospheric layers. Using the fourth-order Runge-Kutta method ( Method) performs numerical integration to calculate the gradient change of atmospheric refractive index; this method can handle areas with more drastic refractive index changes in the atmospheric refractive index field, and help locate areas with sharp refractive index changes, which has an important impact on the path of electromagnetic wave propagation; areas with drastic refractive index gradient changes are the most sensitive parts of electromagnetic wave propagation, which often cause deviations in the signal propagation path; three-dimensional ray tracing equations are used to calculate the actual propagation path of electromagnetic waves emitted from detection stations or satellites; based on the refractive index field and gradient change areas, the actual propagation path of electromagnetic waves is determined, taking into account the influence of factors such as the atmosphere and ionosphere; the differences between these actual propagation paths and ideal paths reflect the propagation errors caused by reasons such as refractive index gradient changes and ionospheric disturbances; by calculating the deviation between the actual propagation path of electromagnetic waves and the apparent path, a geometric correction factor can be obtained; this factor is used to correct the errors caused by atmospheric refraction and other environmental factors Impact on the propagation path of electromagnetic waves; further, the path delay compensation amount is determined based on the geometric correction factor, which can help reduce the time delay error caused by refraction and irregular propagation; ionospheric disturbance parameters are used to describe the impact of the ionosphere on the propagation of electromagnetic waves. Ionospheric disturbances include total electron content (TEC) and scintillation index (S4), which affect the phase of electromagnetic wave propagation; an ionospheric equivalent phase screen model is constructed, and the spatial distribution of phase disturbances is described by a piecewise linear method; this model can simulate the impact of ionospheric disturbances on signals; by analyzing the spatial distribution of phase disturbances, the phase delay and phase scintillation caused by the ionosphere can be calculated, and the phase compensation factor can be further determined based on this information; finally, the path delay compensation amount and the phase compensation factor are vector superimposed to obtain the final propagation path correction factor; this correction factor can be used to correct the propagation path of electromagnetic waves, reduce signal propagation errors, and improve the accuracy of the system.

[0029] In an optional embodiment, an ensemble Kalman filter algorithm is used to construct an assimilation analysis model for the error-corrected target group spatial distribution characteristic data to obtain a data assimilation model, including: D1. Project the background field data of the numerical weather prediction model to the spatiotemporal coordinate system of the detection observation space using a bilinear interpolation algorithm to obtain a spatiotemporally aligned background field dataset. The background field data includes wind field, temperature field, and pressure field. D2. Perform variable conversion on the error-corrected target group spatial distribution characteristic data to obtain an equivalent momentum flux field; generate disturbances through the background field error covariance matrix according to singular value decomposition to obtain an initial disturbance set; D3. Match the disturbance amplitude of the initial disturbance set with the uncertainty magnitude of the background field to obtain the initial state set of the numerical weather prediction model; apply random error disturbance to the detection observation field to obtain the observation set.

[0030] It should be noted that background field data refers to the initial atmospheric data obtained through numerical weather prediction models, usually including wind field, temperature field and pressure field; the background field data is converted from the original coordinate system (such as the three-dimensional grid of the atmosphere) to the space-time coordinate system of the detection observation space through the bilinear interpolation algorithm; the purpose of this conversion is to ensure that during the detection observation process, the background data and the detection observation data are aligned in time and space to ensure the consistency and comparability of the data; by performing variable conversion on the spatial distribution characteristic data of the target group after error correction, a new physical quantity, namely the equivalent momentum flux field, is obtained; this is usually to make the original weather forecast data better match the detection observation data and facilitate the subsequent assimilation process; Singular Value Decomposition (SVD): through singular values Decomposition can extract important disturbance patterns from the background field error covariance matrix; these disturbances represent the uncertainty in the background field; the disturbance set generated by the background field error covariance matrix describes the uncertainty in the initial state; this can provide different initial conditions for weather forecasts to improve the robustness of the forecast; the disturbance amplitude represents the size of the disturbance and usually needs to match the uncertainty magnitude of the background field; this matching can ensure that the disturbance is within a certain range and avoid excessive modification of the background field; to simulate the errors in actual observations, random error disturbances are applied to the detection observation data to generate an observation set; this is to simulate the errors that may occur in the detection observation data in reality, such as noise or systematic errors; the Ensemble Kalman Filter (EnKF) is used for assimilation analysis; its basic idea is to combine multiple possible initial state sets and adjust them through Kalman filtering; by integrating the numerical weather forecast model for each member of the initial state set, the atmospheric state forecast for the future time is obtained; these predicted states are mapped to the detection observation space to obtain the predicted observation set.

[0031] In an optional embodiment, the assimilation analysis model is constructed by using an ensemble Kalman filter algorithm to perform assimilation analysis on the error-corrected spatial distribution characteristic data of the target group to obtain a data assimilation model, further comprising: D4. Perform a numerical weather prediction model integration on each member of the initial state set to predict the atmospheric state at a future time, thereby obtaining a predicted state set; map the predicted state set to the detection observation space, determine the equivalent predicted observation quantity, and obtain a predicted observation set; D5. Determine the covariance matrix of the predicted state set and the covariance matrix of the predicted observation set; determine the correction weights of the observation data on the state of the numerical weather prediction model based on the covariance analysis to obtain the Kalman gain matrix; D6. Dynamically adjust the predicted state set based on the Kalman gain matrix and combine it with the constraint information of the detection observation data to obtain the analysis state set; update the error covariance matrix based on the statistical characteristics of the analysis state set to obtain the data assimilation model.

[0032] It should be noted that the covariance matrix is ​​used to describe the correlation between the predicted state set and the predicted observation set; through covariance analysis, the weight of the detection observation data on the model state correction can be determined, that is, the Kalman gain matrix; the role of the Kalman gain matrix is ​​to guide how to combine the observation data and the predicted state set to optimize the weather forecast; the Kalman gain matrix is ​​used to dynamically adjust the predicted state set, so as to correct the numerical weather forecast in combination with the detection observation data; through this dynamic adjustment, the real atmospheric state can be better approximated; finally, by updating the statistical characteristics of the analysis state set, a new error covariance matrix is ​​obtained; this matrix is ​​used to describe the uncertainty of the model, and ultimately the data assimilation model is obtained, that is, the weather forecast model combined with the detection observation data.

[0033] Example 3, as Figure 2 As shown, the present invention proposes a data assimilation device, which is applicable to a data assimilation method, including: Data acquisition module 1, which is used to perform multi-beam scanning data acquisition based on the detection system to obtain a detection status data set; perform dynamic target detection and tracking based on the detection status data set to obtain detection target motion trajectory data; Feature analysis module 2, which is used to perform spatial distribution feature analysis on the detection target motion trajectory data to obtain a target group spatial distribution feature data set, wherein the target group spatial distribution feature data set includes the spatial density distribution parameters and velocity field gradient characteristics of the target group; Propagation correction module 3 is used to obtain real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; perform electromagnetic propagation effect modeling on the detection status data set based on the real-time meteorological data to obtain a propagation path correction factor; The error compensation module 4 is used to perform observation error compensation on the target group spatial distribution feature data set according to the propagation path correction factor to obtain the target group spatial distribution feature data after error correction; Data assimilation module 5 is used to construct an assimilation analysis model for the error-corrected target group spatial distribution characteristic data through the ensemble Kalman filter algorithm to obtain a data assimilation model; and deploy the data assimilation model to the numerical weather prediction system.

[0034] Example 4, as Figure 3 The present invention proposes a data assimilation device, comprising: a processor 6 and a memory 7; the memory 7 stores computer-executable instructions; the processor 6 executes the computer-executable instructions stored in the memory 7, so that the processor 6 performs a data assimilation method.

[0035] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A data assimilation method, characterized in that: include: Perform multi-beam scanning data acquisition based on the detection system to obtain a detection status data set; Performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data; Performing spatial distribution feature analysis on the detected target motion trajectory data to obtain a target group spatial distribution feature data set, wherein the target group spatial distribution feature data set includes spatial density distribution parameters and velocity field gradient characteristics of the target group; Acquiring real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; performing electromagnetic propagation effect modeling on the detection status dataset based on the real-time meteorological data to obtain a propagation path correction factor; Performing observation error compensation on the target group spatial distribution characteristic data set according to the propagation path correction factor to obtain error-corrected target group spatial distribution characteristic data; constructing an assimilation analysis model for the error-corrected target group spatial distribution characteristic data by using an ensemble Kalman filter algorithm to obtain a data assimilation model; and deploying the data assimilation model to a numerical weather prediction system; The electromagnetic propagation effect modeling is performed on the detection status data set according to the real-time meteorological data to obtain a propagation path correction factor, including: Constructing a three-dimensional atmospheric refractive index field using a Kriging interpolation algorithm based on the atmospheric refractive index profile data, and combining the detection site location and detection range to obtain a spatially continuously distributed refractive index model; Numerical integration is performed according to the fourth-order Runge-Kutta method to obtain a region where the refractive index gradient changes dramatically; and the actual propagation path of the electromagnetic wave in the region where the refractive index gradient changes dramatically is determined according to a three-dimensional ray tracing equation; determining a geometric correction factor according to a deviation between the actual propagation path of the electromagnetic wave and the apparent path, and determining a path delay compensation amount according to the geometric correction factor; constructing an ionospheric equivalent phase screen model according to the ionospheric disturbance parameters, wherein the ionospheric equivalent phase screen model includes a total electron content and a scintillation index, and describes the spatial distribution of the phase disturbance according to a piecewise linear method; determining a phase delay and a phase scintillation caused by the ionosphere according to the spatial distribution of the phase disturbance and the ionosphere equivalent phase screen model, and determining a phase compensation factor according to the phase delay and the phase scintillation; The path delay compensation amount and the phase compensation factor are vectorially superimposed to obtain a propagation path correction factor.

2. A data assimilation method according to claim 1, characterized in that: Dynamic target detection and tracking are performed based on the detection state data set to obtain detection target motion trajectory data, including: performing a time domain convolution operation on the transmission signal and the detection state data set to obtain high-resolution range image data; performing frequency domain filtering on the high-resolution range image data according to a Doppler filter bank to separate the moving target echo from the stationary clutter; Setting a constant false alarm rate detection threshold to perform target point detection on the moving target echo to obtain an initial target point set; Establishing a detection unit spatial index in a polar coordinate system according to the detection beam pointing parameter, and mapping the initial target point trace set to a grid coordinate system of a detection scanning sector according to the detection unit spatial index to obtain a time-space aligned initial target point trace set; A time series alignment model is constructed according to the detection scanning cycle parameters, and an initial target point trace set obtained by different beam pointings is projected into a unified Cartesian coordinate system through coordinate transformation according to the time series alignment model to obtain a time-space aligned point trace set; Initialize the trajectory of the newly appeared target to obtain a candidate trajectory set, wherein the candidate trajectory set includes an unconfirmed trajectory and a first confirmed trajectory.

3. A data assimilation method according to claim 2, characterized in that: Performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data also includes: Performing interactive multi-model filtering on the first confirmed trajectory to obtain a state prediction value of the target, wherein the interactive multi-model filtering adopts a hybrid state transition model of a uniform speed model, a uniform acceleration model, and a coordinated turning model; State prediction and updating are performed based on Kalman filtering, and the spatial coordinates, radial velocity, and signal-to-noise ratio information of the spatiotemporal alignment point trace set are integrated to obtain an optimized state prediction value of the target; The process noise covariance matrix and the measurement noise covariance matrix are dynamically modified according to the target echo intensity to obtain the adaptive filtering parameter set; the trajectory confirmation threshold and deletion time limit are set to obtain the trajectory status label set.

4. A data assimilation method according to claim 3, characterized in that: Performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data also includes: Establishing an association matrix based on the state prediction value of the target appearance and the spatial proximity and Doppler frequency consistency of the spatiotemporal aligned point trace set, and determining an association probability between the spatiotemporal aligned point trace set and the first confirmed trajectory based on the association matrix; Performing a weighted summation of the length of the first confirmed trajectory and the association probability to obtain a trajectory confidence score; screening the first confirmed trajectory according to the trajectory confidence score to obtain a second confirmed trajectory; The spatial coordinate sequence of the second confirmed trajectory is converted into an Earth-centered fixed coordinate system to obtain detection target motion trajectory data.

5. A data assimilation method according to claim 4, characterized in that: Performing spatial distribution feature analysis on the detected target motion trajectory data to obtain a target group spatial distribution feature dataset, including: According to the three-dimensional spatial coordinate sequence of the detection target motion trajectory data, a target point trace is spatially distributed modeled by a kernel density estimation algorithm to obtain a density distribution surface; Identifying local density extreme value areas in the density distribution surface according to a density peak clustering algorithm to obtain distribution area parameters of the target group, wherein the distribution area parameters include peak density, average density, and contour line enclosed area; Performing regional classification on the density distribution surface according to the distribution area parameters of the target group to obtain high-density areas and low-density areas, and performing noise filtering based on neighborhood connectivity on the low-density areas to obtain a set of spatial density distribution parameters; constructing a three-dimensional velocity vector field according to the velocity vector sequence of the detection target motion trajectory data, and determining a velocity gradient tensor of the three-dimensional velocity vector field according to a five-point central difference method; A convolution operation is performed on the velocity gradient tensor according to a Sobel operator to obtain velocity field gradient features, wherein the velocity field gradient features include maximum gradient amplitude, gradient direction variance, and velocity field curvature.

6. A data assimilation method according to claim 5, characterized in that: The error-corrected target group spatial distribution characteristic data is subjected to an assimilation analysis model through an ensemble Kalman filter algorithm to obtain a data assimilation model, including: The background field data of the numerical weather prediction model is projected onto the spatiotemporal coordinate system of the detection observation space according to a bilinear interpolation algorithm to obtain a spatiotemporally aligned background field data set, wherein the background field data includes wind field, temperature field, and pressure field; The error-corrected target group spatial distribution characteristic data is transformed into variables to obtain an equivalent momentum flux field; the perturbation is generated through the background field error covariance matrix according to singular value decomposition to obtain an initial perturbation set; The disturbance amplitude of the initial disturbance set is matched with the uncertainty magnitude of the background field to obtain the initial state set of the numerical weather prediction model; random error disturbance is applied to the detection observation field to obtain the observation set.

7. A data assimilation method according to claim 6, characterized in that: The assimilation analysis model is constructed by using an ensemble Kalman filter algorithm to the error-corrected target group spatial distribution characteristic data to obtain a data assimilation model, further comprising: Performing a numerical weather prediction model integration on each member of the initial state set to predict the atmospheric state at a future time, thereby obtaining a predicted state set; mapping the predicted state set to a detection observation space, determining an equivalent predicted observation quantity, thereby obtaining a predicted observation set; Determining a covariance matrix of the predicted state set and a covariance matrix of the predicted observation set; determining a correction weight of the observation data on the state of the numerical weather prediction model based on covariance analysis to obtain a Kalman gain matrix; The predicted state set is dynamically adjusted according to the Kalman gain matrix, and the constraint information of the detection observation data is combined to obtain the analysis state set; the error covariance matrix is ​​updated according to the statistical characteristics of the analysis state set to obtain the data assimilation model.

8. A data assimilation device, applicable to a data assimilation method according to any one of claims 1 to 7, characterized in that: include: A data acquisition module, wherein the data acquisition module is used to perform multi-beam scanning data acquisition according to the detection system to obtain a detection status data set; Performing dynamic target detection and tracking according to the detection state data set to obtain detection target motion trajectory data; A feature analysis module, wherein the feature analysis module is used to perform spatial distribution feature analysis on the detection target motion trajectory data to obtain a target group spatial distribution feature data set, wherein the target group spatial distribution feature data set includes spatial density distribution parameters and velocity field gradient characteristics of the target group; a propagation correction module, the propagation correction module being configured to acquire real-time meteorological data, wherein the real-time meteorological data includes atmospheric refractive index profile data and ionospheric disturbance parameters; and perform electromagnetic propagation effect modeling on the detection status dataset based on the real-time meteorological data to obtain a propagation path correction factor; an error compensation module, the error compensation module being used to perform observation error compensation on the target group spatial distribution feature data set according to the propagation path correction factor to obtain error-corrected target group spatial distribution feature data; A data assimilation module is used to construct an assimilation analysis model for the error-corrected target group spatial distribution characteristic data through an ensemble Kalman filter algorithm to obtain a data assimilation model; and deploy the data assimilation model to a numerical weather prediction system.

9. A data assimilation device, characterized in that: include: processor and memory; The memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, so that the processor executes a data assimilation method according to any one of claims 1 to 7.

Images