VTS radar-based ship trajectory identification method and system

Abstract

The invention discloses a VTS radar-based ship track identification method and system, and relates to the technical field of ship traffic monitoring, and the method comprises the steps: constructing a continuity equation, marking a divergence abnormal region, generating a preliminary track point set, extracting a complex signal of the preliminary track point set, setting a dynamic threshold value, and marking a nuclear point set with a high interaction amount. Taking each nuclear point as a starting point, tracking a trajectory along a dominant direction, setting a direction offset threshold value, forming a trajectory set, constructing an Euler grid based on the trajectory set, marking a non-physical trajectory segment, correcting the non-physical trajectory segment, and generating a trajectory point set. According to the method, the complex signal is mapped to the energy density field, the analysis of the energy flow vector field is combined, the recognition precision and robustness of the target trajectory are improved, the target trajectory tracking precision is improved by combining the change of the energy flow and trajectory adjustment, and the time consistency analysis and trajectory correction of the Euler grid are carried out, so that the target trajectory tracking precision is improved. And unreasonable track segments are prevented from being reserved.

Description

Technical Field

[0001] The present invention relates to the technical field of ship traffic monitoring, and in particular to a ship trajectory recognition method and system based on VTS radar. Background Art

[0002] With the continuous development of intelligent ports and shipping traffic management technologies, vessel traffic service systems are playing an increasingly important role in ensuring shipping safety and improving navigation efficiency. The VTS system relies on multi-source sensing equipment such as radar, AIS, and video surveillance to monitor ships in the sea area. Among them, radar is the core detection method for all weather and all time periods. It can achieve continuous tracking of targets under conditions such as low visibility and complex weather. VTS radar has made significant progress in detection accuracy, anti-interference capability and signal processing performance. It can output complex signal data containing in-phase components and orthogonal components, providing a basis for high-precision ship trajectory identification.

[0003] Existing ship trajectory recognition technology still has shortcomings. Current methods mostly process radar echoes at the amplitude or distance information level, lacking the ability to map complex signals into energy density fields and combine them with electromagnetic field energy flow vector fields for analysis. This results in an insufficiently detailed depiction of the target's motion direction and energy change trend. The trajectory post-processing stage lacks a correction mechanism based on physical consistency constraints, which easily retains trajectory segments that do not conform to actual motion laws, affecting subsequent traffic situation analysis and command decisions. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a ship trajectory identification method and system based on VTS radar, which solves the problem that the current method of processing radar echoes mostly stays at the amplitude or distance information level, lacks the ability to map complex signals into energy density fields and combine them with electromagnetic field energy flow vector fields for analysis, resulting in insufficiently detailed characterization of target motion direction and energy change trends. The trajectory post-processing stage lacks a correction mechanism based on physical consistency constraints, which easily retains trajectory segments that do not conform to actual motion laws, affecting subsequent traffic situation analysis and command decision-making.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides a method for identifying ship tracks based on VTS radar, comprising the following steps: Collect complex signal data and perform preprocessing, map the normalized complex signal to a rectangular coordinate system, calculate the energy density field of the radar wave, construct a continuity equation, mark areas of divergence anomalies, generate a preliminary trajectory point set, extract the complex signal of the preliminary trajectory point set, calculate the phase angle, perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction, calculate the interaction amount, set a dynamic threshold, and mark the core point set with high interaction amount; Taking each core point as the starting point, the trajectory is tracked along the dominant direction, and the direction offset threshold is set to form a trajectory set. Based on the trajectory set, the tension vector between adjacent points of each trajectory is calculated. Based on the tension vector between adjacent points, the local chain energy is calculated, the total chain energy of each trajectory is summarized, and the trajectory point position is updated to obtain the adjusted trajectory point set; Construct an Euler grid, calculate the time consistency deviation, set the time deviation threshold, mark the non-physical trajectory segments, correct the non-physical trajectory segments, generate a trajectory point set, and build a visualization interface to display the trajectory point set.

[0007] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the normalized complex signal is mapped to a rectangular coordinate system, the energy density field of the radar wave is calculated, the continuity equation is constructed, and the divergence abnormality area is marked, including: Map the normalized complex signal to a rectangular coordinate system, calculate the energy density field of the radar wave, and construct the energy flow vector field based on the normalized complex signal and the energy density field; Based on the energy density field and energy flow vector field, the continuity equation is constructed, the finite difference method is used to solve the continuity equation, the divergence anomaly area is marked, and the preliminary trajectory point set is generated; Extract the complex signal of the preliminary trajectory point set, calculate the phase angle, and perform phase unwrapping. Perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction. Calculate the time gradient of the energy density field. Combine the time gradient and the dominant direction to calculate the interaction amount. Set a dynamic threshold to mark the core point set with high interaction volume.

[0008] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the method of tracking the trajectory along the dominant direction with each core point as the starting point includes: Taking each core point as the starting point, the trajectory is traced along the dominant direction and the mean square error of the phase direction offset is calculated; The direction offset threshold is set using the statistical analysis method. The direction offset threshold is compared with the mean square error of the phase direction offset. If the mean square error of the phase direction offset is greater than the direction offset threshold, the trajectory is determined to be separated. Otherwise, it is determined to be the same trajectory. All separated trajectories are summarized to form a trajectory set.

[0009] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the steps of calculating the tension vector between adjacent points of each trajectory, calculating the local chain energy based on the tension vector between adjacent points, summarizing the total chain energy of each trajectory, and updating the trajectory point position include: Based on the trajectory set, the tension vector between adjacent points of each trajectory is calculated. Based on the tension vector between adjacent points, the local chain energy is calculated and the total chain energy of each trajectory is summarized. Calculate the partial derivative of the total chain energy with respect to the trajectory point and the adaptive step size, update the trajectory point position, use the gradient descent method to set the number of iterations, and obtain the adjusted trajectory point set.

[0010] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the correction of the non-physical trajectory segment to generate a trajectory point set includes: Based on the adjusted trajectory point set, an Euler grid is constructed, the time consistency deviation is calculated, and a time deviation threshold is set using a statistical analysis method. Trajectories with time consistency deviations greater than the time deviation threshold are screened and marked as non-physical trajectory segments, otherwise they are marked as normal trajectory segments. Correction of non-physical trajectory segments; The corrected non-physical trajectory segments and normal trajectory segments are spliced ​​together to generate a trajectory point set.

[0011] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the step of constructing a visual interface to display the trajectory point set includes: Use the visualization tool Matplotlib to build a visualization interface and display the trajectory point set in real time; Users who have passed real-name verification are allowed to view it.

[0012] As a preferred solution of the ship trajectory identification method based on VTS radar of the present invention, the collecting of complex signal data and preprocessing thereof include: The VTS radar system is used to capture electromagnetic echo signals in the target sea area, collect complex signal data, and perform denoising and normalization processing.

[0013] In a second aspect, the present invention provides a ship trajectory identification system based on VTS radar, comprising: Collection and processing module, used for collecting complex signal data and performing preprocessing; The core point updating module is used for calculating a phase angle, performing two-dimensional Fourier transform on the phase angle to extract a dominant direction, calculating an interaction, setting a dynamic threshold, marking a core point set with high interaction, taking each core point as a starting point, tracking a trajectory along the dominant direction, setting a direction offset threshold, forming a trajectory set, calculating a tension vector between adjacent points of each trajectory based on the trajectory set, calculating a local chain energy based on the tension vector between adjacent points, summarizing a total chain energy of each trajectory, updating a trajectory point position, and obtaining an adjusted trajectory point set. The correction module is used for constructing an Euler grid, calculating a time consistency deviation, setting a time deviation threshold, marking a non-physical trajectory segment, correcting the non-physical trajectory segment, and generating the trajectory point set. The display module is used for constructing a visual interface to display the trajectory point set.

[0014] The present application has the following beneficial effects: The present application improves the recognition accuracy and robustness of the target trajectory by mapping the complex signal to the energy density field and combining the analysis of the energy flow vector field, improves the accuracy of the target trajectory tracking by combining the change of the energy flow and the trajectory adjustment, and avoids the reservation of unreasonable trajectory segments by the time consistency analysis of the Euler grid and the trajectory correction. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 The figure is a running flowchart of the ship trajectory recognition method based on the VTS radar in embodiment 1.

[0016] Figure 2 The figure is a structure schematic diagram of the ship trajectory recognition system based on the VTS radar in embodiment 1. DETAILED DESCRIPTION

[0017] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0018] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the concept of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0019] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment.

[0020] Embodiment 1, refer to Figure 1, which is the first embodiment of the present invention, provides a ship trajectory identification method based on VTS radar, comprising the following steps: S1. Collect complex signal data and perform preprocessing. Map the normalized complex signal to a rectangular coordinate system, calculate the energy density field of the radar wave, construct a continuity equation, mark areas of divergence anomalies, generate a preliminary trajectory point set, extract the complex signal of the preliminary trajectory point set, calculate the phase angle, perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction, calculate the interaction amount, set a dynamic threshold, and mark the core point set with high interaction amount. Specifically, complex signal data is collected and preprocessed, including: The electromagnetic echo signal in the target sea area is captured by the VTS radar system, complex signal data is collected, and denoising and normalization processing are performed. The formula is: E(R,θ,t)=I(R,θ,t)+jQ(R,θ,t), Where E(R,θ,t) is a complex signal representing the radar echo at distance R, azimuth θ, and time t. I(R,θ,t) is the in-phase component representing the real part of the echo signal. Q(R,θ,t) is the quadrature component representing the imaginary part of the echo signal. R is the distance from the radar to the target (ship), θ is the azimuth of the radar scan, t is the timestamp, and j is the imaginary unit.

[0021] Through the normalization and denoising of complex signal data, the system can obtain higher-quality signal input, which helps to accurately identify the position and dynamics of ship targets, effectively avoid the error problems that occur in traditional signal processing, and improve the stability and robustness of the radar system in complex environments.

[0022] Furthermore, the normalized complex signal is mapped to a rectangular coordinate system, and the energy density field of the radar wave is calculated, the continuity equation is constructed, and the divergence anomaly area is marked, including: Map the normalized complex signal to the rectangular coordinate system and calculate the energy density field of the radar wave. The formula is: x=Rcosθ, y=Rsinθ, W(x,y,t)=|E'(x,y,t)| 2 / Z eff , Where W(x,y,t) is the energy density field of the radar wave, which represents the energy intensity at the rectangular coordinate (x,y) and time t, x and y are the rectangular coordinate positions, E'(x,y,t) is the normalized complex signal, and Z eff is the effective impedance, set using the fixed value method, and W(x,y,t) is the energy density field in spacetime; Based on the normalized complex signal and energy density field, the energy flow vector field is constructed, and the formula is: , in is the energy flow vector field, which indicates the direction and intensity of electromagnetic wave energy propagation, μ0 is the vacuum magnetic permeability, which is set based on electromagnetic theory, Re is the real part of the complex number, × is the vector cross product, which indicates the directional propagation of electromagnetic field energy flow, and W max (t) is the maximum value of the energy density field at the current time, ρ eff is the free space wave impedance; Based on the energy density field and energy flow vector field, the continuity equation is constructed, and the formula is: , in is the partial derivative of energy density with respect to time, which indicates the rate of change of energy with time. is the gradient operator, which represents the operator symbol of spatial partial derivatives. is the divergence of the energy flow vector field, Q env is the environmental interference term, which represents the disturbance of energy flow caused by external factors; The finite difference method is used to solve the continuity equation, mark the divergence anomaly area, and generate a preliminary trajectory point set. The formula is: , σ th =μ div +k σ ·σ div , Among them, P init is the preliminary trajectory point set, i is the index of a single trajectory point in the preliminary trajectory point set, σ th is the divergence threshold, which is set using statistical analysis, N is the total number of trajectory points, μ div and σ div are the mean and standard deviation of the divergence, k σ is the dynamic adjustment coefficient of the divergence threshold; Extract the complex signal of the preliminary trajectory point set, calculate the phase angle, and perform phase unwrapping. The formula is: , Where Φ(x,y,t) is the phase angle, which represents the phase angle at the rectangular coordinate (x,y) and time t, unwrap is the phase unwrapping operation, Q'(x,y,t) and I'(x,y,t) are the quadrature component and in-phase component of the normalized complex signal in the preliminary trajectory point set, respectively; Perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction. The formula is: , , in U and V are the frequency coordinates in Fourier domain, N' is the size of Fourier transform window, is the dominant direction vector, representing the local motion direction, is the frequency coordinate corresponding to the maximum value; The time gradient of the energy density field is calculated, and the formula is: , where is the time gradient of the energy density field, and Δt is the radar scanning period; The interaction is calculated by combining the time gradient and the dominant direction, and the formula is: , where is the interaction, representing the dot product of the direction field and the energy gradient, v max is the maximum speed of the ship, obtained by the sensor; The dynamic threshold is set, and the core point set with high interaction is marked, representing the position of the target ship, and the formula is: , , where is the core point set, containing the coordinates and time of the target point, is the dynamic threshold at time t, set by statistical analysis, and are the mean and standard deviation of the interaction, respectively, k' σ is the dynamic adjustment coefficient of the dynamic threshold.

[0023] By calculating the energy density field, the propagation characteristics of the signal can be intuitively presented from the spatial angle, helping to further improve the target recognition ability of the radar signal in complex sea environment. Through the marking of the divergence anomaly area, the credibility and accuracy of the ship trajectory can be significantly improved. The motion direction of the ship is extracted to avoid the problem of unstable target recognition caused by sea surface disturbance, ship maneuverability change and other factors. The combination of interaction and dynamic threshold makes the system flexible to respond to different sea traffic conditions and has high adaptability. By marking the core point set with high interaction, not only the accuracy of ship trajectory recognition is improved, but also the sensitivity to ship interaction is significantly enhanced. Especially in complex environments such as dense port navigation area or bad weather conditions, the target information can be updated in time to realize dynamic monitoring of the ship.

[0024] S2, tracking a trajectory along a dominant direction starting from each core point, setting a direction deviation threshold, forming a trajectory set, calculating a tension vector between adjacent points of each trajectory based on the trajectory set, calculating a local chain energy based on the tension vector between adjacent points, summarizing a total chain energy of each trajectory, updating a trajectory point position, and obtaining an adjusted trajectory point set; Specifically, tracking a trajectory along a dominant direction starting from each core point includes: Tracking a trajectory along a dominant direction starting from each core point has a formula as follows: , where Γ ο (t) is a position of an oth trajectory at time t, v o is a ship speed of the oth trajectory, is a direction vector of a trajectory point at a previous time, and o is an index of a single trajectory; Calculating a phase direction deviation mean square error has a formula as follows: , where Δ Φ (o, u) is a phase direction deviation mean square error of an oth trajectory and an u th trajectory, and are direction vectors of the oth trajectory and the u th trajectory at time t, respectively, and T is a length of a time window; Using a statistical analysis method to set a direction deviation threshold, comparing the direction deviation threshold with the phase direction deviation mean square error, if the phase direction deviation mean square error is greater than the direction deviation threshold, determining that the trajectories are separated, otherwise determining that the trajectories are the same, and summarizing all separated trajectories to form a trajectory set.

[0025] Through dominant direction tracking, the system can ensure that the motion trajectory of a target ship can be accurately predicted when the direction of the ship changes, avoiding the delay response or deviation of the trajectory change in the traditional method, and through intelligent trajectory separation, the target recognition and tracking efficiency in a multi-target scene can be improved, especially in a complex navigation area (such as a port or a channel intersection), the motion path of each ship can be more accurately identified, and misidentification can be prevented.

[0026] Further, calculating a tension vector between adjacent points of each trajectory based on the trajectory set includes: Calculating a tension vector between adjacent points of each trajectory based on the trajectory set has a formula as follows: , where is a tension vector of a u th trajectory, u = (x u , y u,t u ) is the position and time of the u-th trajectory point; Based on the tension vector between adjacent points, the local chain energy is calculated and the total chain energy of each trajectory is summarized as follows: , , , Among them E u is the local chain energy of the u-th trajectory, t u is the timestamp of the u-th trajectory point, v u is the instantaneous velocity of the u-th trajectory, σ u is the standard deviation of velocity, E' is the total chain energy; Calculate the partial derivative of the total chain energy with respect to the trajectory point and the adaptive step size, the formula is: , , in is the partial derivative of the total chain energy with respect to the trajectory point, γ is the gradient weight coefficient, which is set using the linear search method, η is the adjustment step size, which is set using the linear weighting method, and α η and β η are the step size ratio coefficient and the environmental interference adjustment coefficient, respectively, and are set using grid search optimization. env is the environmental interference intensity, set using empirical rules; Update the position of the new trajectory point, use the gradient descent method to set the number of iterations, and get the adjusted trajectory point set. The formula is: , Among them, P u new is the u-th trajectory point after update.

[0027] The introduction of tension vector can enhance the description of the relationship between trajectory points and help the system accurately calculate the instantaneous speed change of the ship. By calculating the local chain energy, the system can effectively evaluate whether each trajectory point conforms to the physical motion law of the ship, and in the trajectory update process, it can avoid overfitting or false trajectory changes, thereby improving the stability and accuracy of trajectory prediction. By optimizing the partial derivatives of the total chain energy with respect to the trajectory points, the position and motion trajectory of the trajectory points can be finely adjusted. The introduction of the gradient descent method enables the system to obtain accurate trajectory point positions through step-by-step optimization, ensuring the dynamic update and optimization of the trajectory point set. In a multi-target environment, iterative optimization can ensure that the system accurately tracks the changes of each target, especially in the case of high-speed movement or complex trajectory changes, and can quickly adjust and reduce errors, thereby improving the accuracy of target recognition.

[0028] S3. Construct an Euler grid, calculate the time consistency deviation, set the time deviation threshold, mark the non-physical trajectory segments, correct the non-physical trajectory segments, generate a trajectory point set, and build a visualization interface to display the trajectory point set; Specifically, the non-physical trajectory segments are corrected to generate a trajectory point set, including: Based on the adjusted trajectory point set, the Euler grid is constructed and the time consistency deviation is calculated. The formula is: , , where δ s is the temporal consistency deviation of the sth grid point, M is the neighborhood of the grid point The number of trajectory points within is the neighborhood of the sth grid point, set by the spatial neighborhood partitioning method, μ flow is the global time change rate mean, P ' u and P ' u+1 is a pair of trajectory points in the neighborhood; Use statistical analysis to set a time deviation threshold, filter out trajectories with time consistency deviations greater than the time deviation threshold, and mark them as non-physical trajectory segments; otherwise, mark them as normal trajectory segments. Correct the non-physical trajectory segment using the following formula: , , in is the corrected direction vector of the uth segment, β' is the historical velocity weight, which is set using the linear search method. is the historical speed average, P * u is the corrected trajectory point; The corrected non-physical trajectory segments and normal trajectory segments are spliced ​​together to generate a trajectory point set.

[0029] By constructing an Euler grid, the system can uniformly map the spatial distribution and temporal variations of trajectory points onto the grid, providing more accurate trajectory analysis. The calculation of temporal consistency deviation can promptly identify non-physical components of the trajectory, ensuring that the system only uses valid data that conforms to physical laws, thereby improving the reliability of trajectory prediction and analysis. The temporal deviation threshold, set through statistical analysis, can be dynamically adjusted to suit different environments and application scenarios, providing high flexibility. By eliminating or correcting non-physical trajectory segments, the system can eliminate erroneous trajectories caused by sensor errors, data loss, or interference, improving the authenticity and reliability of trajectory data and ultimately enhancing the accuracy of target identification and tracking. A correction algorithm based on historical velocity weights can more accurately restore the true motion path of non-physical trajectory segments. A combination of a linear search method and the historical velocity mean enables intelligent trajectory correction, automatically adapting to the motion characteristics of different ships. Through correction and splicing, the system ensures that the generated trajectory point set is uninterrupted and reasonably reflects the ship's true motion path. By splicing the corrected trajectory segments, the system can effectively integrate trajectory data from different time periods, providing comprehensive target trajectory information.

[0030] Furthermore, a visualization interface is constructed to display the trajectory point set, including: Use the visualization tool Matplotlib to build a visualization interface and display the trajectory point set in real time; Users who have passed real-name verification are allowed to view it.

[0031] Through the visual interface, users can intuitively view the track point set and its changes, providing a friendly operation interface, improving the readability and usability of the data, and allowing users with real-name verification to access track data. Not only can they view ship tracks in real time, but they can also perform data query and analysis operations, providing personalized data display and analysis services for different users.

[0032] Example 2, reference Figure 2 In a second embodiment of the present invention, a ship trajectory identification system based on VTS radar includes: Collection and processing module, used for collecting complex signal data and performing preprocessing; The kernel point update module is used to calculate the phase angle, perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction, calculate the interaction amount, set a dynamic threshold, mark the kernel point set with high interaction amount, use each kernel point as the starting point, track the trajectory along the dominant direction, set the direction offset threshold, form a trajectory set, calculate the tension vector between adjacent points of each trajectory based on the trajectory set, calculate the local chain energy based on the tension vector between adjacent points, summarize the total chain energy of each trajectory, update the trajectory point position, and obtain the adjusted trajectory point set; The correction module is used to construct the Euler grid, calculate the time consistency deviation, set the time deviation threshold, mark the non-physical trajectory segments, correct the non-physical trajectory segments, and generate the trajectory point set; The display module is used to build a visual interface to display the trajectory point set.

[0033] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A ship trajectory recognition method based on VTS radar, characterized by: The steps include: Collect complex signal data and perform preprocessing. Map the normalized complex signal to a rectangular coordinate system, calculate the energy density field of the radar wave, construct a continuity equation, mark areas of divergence anomalies, generate a preliminary trajectory point set, extract the complex signal from the preliminary trajectory point set, calculate the phase angle, perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction, calculate the interaction amount, set a dynamic threshold, and mark the core point set with high interaction amount. Taking each core point as the starting point, the trajectory is tracked along the dominant direction, and the direction offset threshold is set to form a trajectory set. Based on the trajectory set, the tension vector between adjacent points of each trajectory is calculated. Based on the tension vector between adjacent points, the local chain energy is calculated, the total chain energy of each trajectory is summarized, and the trajectory point position is updated to obtain the adjusted trajectory point set; Construct an Euler grid, calculate the time consistency deviation, set the time deviation threshold, mark the non-physical trajectory segments, correct the non-physical trajectory segments, generate a trajectory point set, and build a visualization interface to display the trajectory point set.

2. The ship trajectory identification method based on VTS radar according to claim 1, characterized in that: The normalized complex signal is mapped to a rectangular coordinate system, and the energy density field of the radar wave is calculated, the continuity equation is constructed, and the divergence anomaly area is marked, including: Map the normalized complex signal to a rectangular coordinate system, calculate the energy density field of the radar wave, and construct the energy flow vector field based on the normalized complex signal and the energy density field; Based on the energy density field and energy flow vector field, the continuity equation is constructed, the finite difference method is used to solve the continuity equation, the divergence anomaly area is marked, and the preliminary trajectory point set is generated; Extract the complex signal of the preliminary trajectory point set, calculate the phase angle, and perform phase unwrapping. Perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction. Calculate the time gradient of the energy density field. Combine the time gradient and the dominant direction to calculate the interaction amount. Set a dynamic threshold to mark the core point set with high interaction volume.

3. The ship trajectory identification method based on VTS radar according to claim 2, characterized in that: The method of tracking the trajectory along the dominant direction with each core point as the starting point includes: Taking each core point as the starting point, the trajectory is traced along the dominant direction and the mean square error of the phase direction offset is calculated; The direction offset threshold is set using the statistical analysis method. The direction offset threshold is compared with the mean square error of the phase direction offset. If the mean square error of the phase direction offset is greater than the direction offset threshold, the trajectory is determined to be separated. Otherwise, it is determined to be the same trajectory. All separated trajectories are summarized to form a trajectory set.

4. The ship trajectory identification method based on VTS radar according to claim 3, characterized in that: The calculation of the tension vector between adjacent points of each trajectory, the calculation of the local chain energy based on the tension vector between adjacent points, the aggregation of the total chain energy of each trajectory, and the update of the trajectory point position include: Based on the trajectory set, the tension vector between adjacent points of each trajectory is calculated. Based on the tension vector between adjacent points, the local chain energy is calculated and the total chain energy of each trajectory is summarized. Calculate the partial derivative of the total chain energy with respect to the trajectory point and the adaptive step size, update the trajectory point position, use the gradient descent method to set the number of iterations, and obtain the adjusted trajectory point set.

5. The ship trajectory identification method based on VTS radar according to claim 4, characterized in that: The step of correcting the non-physical trajectory segment to generate a trajectory point set includes: Based on the adjusted trajectory point set, an Euler grid is constructed, the time consistency deviation is calculated, and a time deviation threshold is set using a statistical analysis method. Trajectories with time consistency deviations greater than the time deviation threshold are screened and marked as non-physical trajectory segments, otherwise they are marked as normal trajectory segments. Correction of non-physical trajectory segments; The corrected non-physical trajectory segments and normal trajectory segments are spliced ​​together to generate a trajectory point set.

6. The ship trajectory identification method based on VTS radar according to claim 5, characterized in that: The construction of a visualization interface to display the trajectory point set includes: Use the visualization tool Matplotlib to build a visualization interface and display the trajectory point set in real time; Users who have passed real-name verification are allowed to view it.

7. The method for ship trajectory recognition based on VTS radar according to claim 6, characterized in that: The collecting of complex signal data and preprocessing thereof include: The VTS radar system is used to capture electromagnetic echo signals in the target sea area, collect complex signal data, and perform denoising and normalization processing.

8. A ship trajectory identification system based on VTS radar, used to implement the method according to any one of claims 1 to 7, characterized in that: include: Collection and processing module, used for collecting complex signal data and performing preprocessing; The kernel point update module is used to calculate the phase angle, perform a two-dimensional Fourier transform on the phase angle to extract the dominant direction, calculate the interaction amount, set a dynamic threshold, mark the kernel point set with high interaction amount, use each kernel point as the starting point, track the trajectory along the dominant direction, set the direction offset threshold, form a trajectory set, calculate the tension vector between adjacent points of each trajectory based on the trajectory set, calculate the local chain energy based on the tension vector between adjacent points, summarize the total chain energy of each trajectory, update the trajectory point position, and obtain the adjusted trajectory point set; The correction module is used to construct the Euler grid, calculate the time consistency deviation, set the time deviation threshold, mark the non-physical trajectory segments, correct the non-physical trajectory segments, and generate the trajectory point set; The display module is used to build a visual interface to display the trajectory point set.

Images