53 results about "Ship tracks" patented technology

The invention discloses a near-shore ship tracking method and device based on multi-source data fusion and a storage medium, and the method comprises the steps: obtaining and processing first-class radar point cloud data of a wind power plant millimeter wave radar, vibration sensor data of a breakwater, AIS data and second-class radar point cloud data of multiple radar stations; performing space alignment and time alignment on the processed first type of radar point cloud data and the processed second type of radar point cloud data, and performing time alignment on the processed AIS data; fusing the aligned data to obtain target state information; and predicting a plurality of states of the ship at future moments based on the target state information to obtain a plurality of corresponding prediction trajectories, correcting the prediction trajectories based on a wind power plant wake flow model, and screening out an optimal trajectory from the corrected prediction trajectories as a ship track. Therefore, the method can improve the multi-dimensional anti-risk capability, has higher safety guarantee, and can improve the supervision efficiency in a novel offshore wind power plant and other complex environments.

The invention discloses a visual drawing method and device for an inland ship navigation trajectory and a medium, and the method comprises the steps: collecting GPS positioning data, AIS dynamic parameters and static attribute data of a ship through a ship terminal, carrying out the preprocessing of the GPS positioning data and the AIS dynamic parameters, and storing the processed data in a database according to a timestamp index; calling a geocoding service interface, mapping the processed longitude and latitude coordinates of the ship to an electronic navigation channel chart coordinate system, and periodically updating a ship position mark in an electronic map interface; in response to a time period query instruction input by a user, extracting ship track point data corresponding to the timestamp sequence from the database, and superposing the ship track point data to the electronic channel chart to generate a dynamic track line; a user operation instruction is received through the time axis control, the ship mark position and the course display state on the trajectory line are dynamically adjusted, and visualized backtracking of the historical trajectory of the ship is achieved.

The invention discloses a moored ship motion state prediction and anti-collision early warning system, particularly relates to the field of traffic control and anti-collision early warning, and is used for solving the problems that ship motion state accurate prediction and anti-collision scheduling response lag cannot be realized in existing port mooring operation. According to the system, ship trajectory, port operation and environment monitoring data are acquired in real time through a data acquisition module, and a fusion data stream with consistent time-space attributes is generated; constructing a space-time diagram structure by using a feature extraction module, extracting traffic flow dynamic features, and calculating a short-term conflict probability, a trajectory convergence risk value and a mooring risk value; the risk early warning module generates a collision risk index based on risk level mapping and triggers graded early warning; the mooring decision-making module sets a key control leg according to the early warning level and generates a dynamic decision-making scheme; the anti-collision scheduling module executes actual channel passing, speed adjustment and track change, and dynamic coordination and safety management and control of port mooring anti-collision are achieved.

The embodiment of the invention provides a leaning behavior recognition method and device, electronic equipment, a readable storage medium and a chip, and the method comprises the steps: obtaining a first driving track of double ships and corresponding first track data; determining a leaning behavior track section based on the first traveling track and the first track data of the double ships; determining a single ship trajectory positive sample data set according to the lapping behavior trajectory segment; determining a single ship trajectory negative sample data set according to the normal navigation trajectory segment; training a fusion neural network recognition model according to the single-ship trajectory positive sample data set and the single-ship trajectory negative sample data set; obtaining target trajectory data of a to-be-identified single target ship; determining a track fragment corresponding to the target track data; and determining a leaning behavior state corresponding to at least one track fragment of the single target ship. According to the scheme of the invention, under the condition that only one ship starts the AIS, the leaning behavior is identified, and the completeness of maritime behavior supervision is improved.

The invention discloses a ship dynamic high-precision positioning and abnormal behavior recognition method, and particularly relates to the technical field of navigation positioning. The method comprises the following steps: performing time sequence synchronization and noise elimination on satellite signal propagation time ranging data, carrier phase data and ship attitude data received by a ship; outputting a reflection path coupling relation parameter by analyzing a time-varying coupling relation between the rolling and pitching states of the ship body and a sea surface specular reflection path; by establishing a multi-path deviation correction model, error correction values of satellite signal propagation time ranging data and carrier phase data are calculated; ship track motion state data are output by extracting ship space attitude change features and track geometric morphology features; and updating the ship trajectory motion state data by using the corrected satellite signal propagation time ranging data and carrier phase data to obtain ship positioning trajectory data, and outputting a ship abnormal behavior identification result. And the detection accuracy and reliability of the abnormal behaviors of the ship are improved.

This invention discloses a computer vision-based method for ship traffic situational awareness, comprising: collecting ship traffic data and preprocessing it to obtain a video frame sequence; performing ship target detection and outputting a set of ship detection results; constructing an improved ByteTrack model and outputting a set of trajectory segments; performing a Poisson-Dobernoli hybrid filter to obtain a set of ship tracks; performing coordinate mapping and track element calculation, calculating speed, heading, and position; performing risk event determination and outputting risk event results; and performing graded early warning and interpretable processing to generate early warning information. This invention, through the collaborative processing of the improved ByteTrack model and the Poisson-Dobernoli hybrid filter, achieves graded early warning and interpretable output of risk events for ship targets under complex sea surface imaging conditions.

The application discloses a near-shore ship tracking method and device based on multi-source data fusion and a storage medium. The method comprises the following steps: acquiring first radar point cloud data of a wind farm millimeter wave radar, breakwater vibration sensor data, AIS data and second radar point cloud data of a multi-radar station, and processing the data; performing spatial alignment and time alignment on the processed first radar point cloud data and second radar point cloud data, and performing time alignment on the processed AIS data; fusing the aligned data to obtain target state information; predicting multiple states of a ship at a future time based on the target state information, obtaining corresponding multiple predicted trajectories, correcting the predicted trajectories based on a wind farm wake model, and screening an optimal trajectory from the corrected predicted trajectories as a ship track. Thus, the multi-dimensional risk resistance capability can be improved, the safety guarantee is higher, and the supervision efficiency in a complex environment such as a new near-shore wind farm can be improved.

The invention discloses a navigation mark visual identification and collision early warning method and system based on deep learning, and relates to the technical field of intelligent shipping, and the method comprises the steps: collecting a navigation channel monitoring video stream in real time, and carrying out the preprocessing of a key image frame; constructing a navigation mark detection model based on an improved Officient Det network, and outputting the position and category of the navigation mark in the key image frame; the method comprises the following steps: constructing a ship track prediction model through ship historical track data, predicting a short-time track of a ship, calculating the relative position, speed and course angle of the ship and a navigation mark, and constructing a dynamic collision risk field; and fusing the visual detection result, the radar ranging data and the AIS information to obtain a ship collision risk, and triggering graded early warning based on a preset risk threshold. According to the method, the improved deep learning model and the real-time calculation framework are combined, the robustness of navigation mark identification is improved, an efficient collision early warning mechanism is established, and technical support is provided for intelligent shipping.

The utility model discloses a ship trackball with a drainage structure, which comprises a shell, a top cover, a bottom plate and a trackball body, the top cover and the bottom plate are respectively fixed on the upper surface and the lower surface of the shell, the trackball body is movably arranged in the shell through a plurality of stainless steel bearings, and a dustproof ring is arranged between the top cover and the trackball body. The surface of the dustproof ring is sleeved with high-elasticity foam relative to the bottom of the top cover. A drainage structure is arranged at the bottom of the bottom plate; according to the ship trackball, an existing ship trackball is improved, the drainage structure is designed at the bottom of the ship trackball, water in equipment can be smoothly drained from the bottom under the condition that the bottom plate is not detached, and due to the design, the waterproof performance of the equipment is improved, the maintenance process is simplified, the use cost of a user is reduced, and the service life of the ship trackball is prolonged. Meanwhile, the high-elasticity foam is designed at the top, so that a gap between the upper part and the lower part of the ball body can be buffered, and moisture permeating from the top surface can be absorbed, and therefore, the PCB inside is protected, and the overall reliability of equipment is ensured.

The invention provides a marine ship trajectory association method, device and equipment and a medium, and the method comprises the steps: employing OpenCV to extract a marine ship image, carrying out the preprocessing, and employing edge detection and contour processing, thereby obtaining the size features of a marine ship; extracting the contour gravity center position of the marine ship based on a six-parameter affine transformation matrix of GDAL, and extracting longitude and latitude coordinates of a target; obtaining the multi-modal features of the marine ship, carrying out the weighted fusion of the multi-modal features, and taking N targets with the highest probability as a candidate set; based on the dynamic behavior characteristics, obtaining the course and speed of the marine ship and then calculating a predicted target position; and matching the longitude and latitude coordinates of the target in the candidate set with the predicted target position, selecting the predicted position with the minimum distance as an association point through Euclidean distance calculation, generating a continuous track and forming target situation information. According to the technical scheme, the target recognition precision can be improved, and the continuous target situation can be formed.

The invention discloses an intelligent track planning system and method applied to an unmanned ship, and relates to the technical field of track planning, and the method comprises the steps: shooting a target image, setting the moving speed and collection interval of the unmanned ship, and obtaining an initial navigation track; historical navigation records of the unmanned ship are called, and a reference track in the navigation tracks is judged and extracted; acquiring a reference image of the water area, and extracting an obstacle area in the to-be-detected water area; and extracting a plurality of sub-trajectories in the to-be-detected trajectory to obtain all collection points, setting index conditions, performing early warning prompt if the index conditions are met, and changing the to-be-detected trajectory again until the adjusted navigation trajectory does not meet the index conditions. According to the invention, by analyzing the historical navigation record, judging the rationality of the adjusted navigation track and carrying out early warning prompt, intelligent planning of the unmanned ship track is realized, and the unmanned ship track planning capability and the measured water depth level are effectively improved.

The invention discloses a ship operation data backtracking method based on an error correction algorithm, which relates to the field of maritime engineering, solves the problem of high error rate of ship operation data abnormity backtracking, and comprises the following steps: setting a track monitoring area of a target ship in the navigation process, analyzing ship navigation data based on the track monitoring area, and determining the ship operation data according to the analyzed ship navigation data. Obtaining an optimized ship track of the target ship; the optimized ship track is analyzed, the target ship is classified according to the historical track of the historical ship, and the historical ship of the same category as the target ship is obtained; analyzing the ship operation data, judging an abnormal condition of the ship operation data corresponding to the target ship in combination with historical operation data, and generating a corresponding abnormal signal; after the abnormal signal is received, a ship operation function is constructed according to the historical operation data of the same type of historical ships, abnormity backtracking is conducted on the ship operation data of the target ship based on the ship operation function, and accurate backtracking of the ship operation data is achieved.

The invention discloses a ship platform abnormal behavior detection method based on ADS-B big data, and particularly relates to the technical field of ship navigation safety monitoring. The method comprises the following steps: acquiring automatic related monitoring broadcast data of a target sea area ship, constructing a ship track evolution matrix, and generating a ship perturbation motion feature tensor; constructing a ship neighborhood interactive topology network, calculating a trajectory disturbance propagation weight between nodes, and forming an abnormal propagation probability matrix; establishing a coupling response relation between a local track disturbance energy sequence and neighborhood interaction intensity through time sequence folding mapping, generating an abnormal behavior evolution prediction curve, calculating an abnormal energy aggregation index, generating a ship abnormal behavior risk distribution map in combination with an abnormal propagation probability matrix, extracting implicit behavior indexes, and calculating a ship abnormal behavior risk distribution map. Through dynamic matching with the preset abnormal behavior mode, the micro-scale abnormal navigation behavior in a complex sea area environment can be effectively identified, and the accuracy and the early warning capability of ship abnormal behavior detection are improved.

The application provides a ship voyage track generation method based on STK and MATLAB combined simulation, comprising: setting a random seed, initializing an STK environment through MATLAB, and loading a set simulation scene; under the simulation scene, randomly selecting a ship model needing to be simulated, and setting a corresponding type, attribute and speed for each ship model; using a random algorithm to generate a starting point and an ending point of each ship model, using an interpolation method to generate continuous and diverse path points, and through dynamic propagation of the path points, updating a voyage track of the ship model in real time. The application not only improves the efficiency and practicability of the voyage track simulation, but also greatly improves the convenience and accuracy of the voyage track data analysis through automatic data processing and output functions, and is suitable for fields of ocean research, navigation safety evaluation, channel planning, military and education training and the like.

The application discloses a kind of ship dynamic high-precision positioning and abnormal behavior identification method, specifically related to navigation positioning technical field;It is through satellite signal propagation time ranging data received by ship, carrier phase data and ship attitude data are time series synchronized and noise is removed;Through the time-varying coupling relationship of the analysis ship body roll and pitch state and sea surface mirror reflection path, output reflection path coupling relationship parameter;Through establishing multipath deviation correction model, the error correction value of satellite signal propagation time ranging data and carrier phase data is calculated;Through extracting ship spatial attitude change characteristics and track geometric shape characteristics, output ship track motion state data;The satellite signal propagation time ranging data and carrier phase data after correction are used to update ship track motion state data, obtain ship positioning track data, output ship abnormal behavior identification result;It improves the detection accuracy and reliability of ship abnormal behavior.

The invention provides an LSTM-based false AIS ship trajectory detection method, device and equipment and a storage medium, deep learning is performed on a ship historical behavior mode through an LSTM neural network, behavior characteristics are extracted from multiple dimensions such as speed, course and trajectory form, a time sequence dependency relationship is captured by using a memory mechanism of the LSTM, and a personalized'behavior fingerprint 'is established for each ship. And through a normal behavior index database constructed by KDTree, dynamic threshold detection based on a historical mode of the ship is realized, and a high false alarm rate caused by a unified threshold is avoided. During real-time detection, similarity matching is carried out on the current track characteristics and historical behavior fingerprints, so that not only can obvious position jump and other abnormalities be identified, but also careful forged tracks deviating from an inherent behavior mode can be found, and a technical breakthrough from single-dimensional rule detection to multi-dimensional intelligent identification is realized.

The application provides a mobile ship tracking task optimization method and system based on an ADMM algorithm, relates to the technical field of satellite task planning, and solves the technical problems of high calculation complexity, dynamic response lag, and observation blind area caused by multi-satellite resource competition in the prior art. The method comprises the following steps: acquiring a target sea area image to be observed; dividing the target sea area image into a plurality of grid cells; setting an observation time period and dividing the observation time period into a plurality of observation sub-periods; acquiring a plurality of visible satellites corresponding to the observation sub-periods, assigning grid cells to each visible satellite, and obtaining a corresponding observation ship list; initializing algorithm parameters; optimizing and detecting conflicts of the plurality of observation ship lists through the ADMM algorithm to obtain an optimal observation list; and observing and sorting the plurality of ships in the optimal observation list in chronological order to obtain an observation task plan table. The application is used in the process of tracking mobile ships at sea.

The invention provides a mobile ship tracking task optimization method and system based on an ADMM algorithm, relates to the technical field of satellite task planning, and solves the technical problems of high calculation complexity, dynamic response lag and observation blind areas caused by multi-satellite resource competition in the prior art. The method comprises the following steps: acquiring a to-be-observed target sea area image; dividing the target sea area image into a plurality of grid units; setting an observation time period, and dividing the observation time period into a plurality of observation sub-periods; obtaining a plurality of corresponding visual satellites based on the observation sub-cycles, and distributing grid units and corresponding observation ship lists for each visual satellite; initializing algorithm parameters; performing optimization and conflict detection on the plurality of observation ship lists through an ADMM algorithm to obtain an optimal observation list; and carrying out observation sorting on the plurality of ships in the optimal observation list according to a time sequence to obtain an observation task schedule. The method is used in the marine mobile ship tracking process.

The application provides a tracking control method for a full-drive water surface ship with time-varying control gain, comprising: establishing a full-drive water surface ship system model considering external disturbance; designing a tracking controller combined with time-varying control gain and non-time-varying control gain by using vector backstepping method; analyzing asymptotic tracking performance and practical tracking performance of the tracking controller by using Lyapunov stability theory; analyzing boundedness of control items containing time-varying control gain and control items containing non-time-varying control gain in the tracking controller, and realizing tracking control of the full-drive water surface ship. The application designs a tracking control method containing time-varying control gain based on unknown bounded external disturbance, which is used to realize asymptotic tracking performance of the system and stability of the closed loop system; on the other hand, only by adjusting the value of the time-varying control gain, the asymptotic tracking or practical tracking closed loop performance of the system can be realized.

This invention discloses an active collision avoidance method for offshore wind turbine monopile foundations, aiming to provide a method that combines the advantages of active collision avoidance early warning with passive collision avoidance, thus providing better protection for offshore wind turbine monopile foundations. Potential collision objects are classified and identified: if the potential collision object is not a ship, a collision limit safety zone S1 is defined within the monitoring range; if the potential collision object enters the collision limit safety zone S1, the water level is determined and a collision avoidance device is deployed. If the potential collision object is a ship, the collision limit safety zone S1, a key monitoring area S2, and a tracking and identification area S3 are defined within the monitoring range. The ship's trajectory is predicted, and a risk assessment of the collision between the ship and the wind turbine monopile foundation is performed. A collision early warning threshold is generated, and the risk assessment result is compared with the collision early warning threshold; it is determined whether the collision early warning threshold is exceeded; if the threshold is exceeded, an alarm is issued and monitoring continues.

The patent belongs to the field of remote sensing data marine application processing, and particularly relates to a marine ship correlation identification method based on AIS and satellite-borne SAR data. The method mainly comprises the following steps: step one: acquiring AIS data based on a SAR satellite imaging time; step two: AIS data cleaning and processing; step three: ship track recovery in an AIS time interval; step four: ship state deduction in the AIS time interval; the deduction comprises time-distance vector or time-azimuth vector speed mapping relationship establishment; step five: ship real position revision based on a satellite-borne SAR imaging principle; step six: AIS and satellite-borne SAR data marine ship target accurate correlation processing; and step seven: satellite-borne SAR marine image system geometric error calculation. The method deduces ship real position, speed, heading and other state information at the satellite imaging time through data processing, mapping relationship establishment, distance speed decomposition and other modes, and can effectively quantify the degree of displacement of a moving target when imaged by a SAR.

The invention discloses a ship track keeping guidance law design method based on a vector field. The method comprises the following steps: establishing a three-degree-of-freedom kinematics model of a ship; based on a transverse error and a longitudinal error between the ship and the target path of the three-degree-of-freedom kinematics model, deriving the transverse error and the longitudinal error to obtain dynamic characteristics of the errors; then, through the dynamic characteristics of the transverse error and the longitudinal error and according to the Lyapunov stability theory, a ship track preserving vector field is constructed, and the ship track preserving vector field comprises two components; and finally, converting the component of the ship track keeping vector field into an expected course angle and an expected speed instruction. According to the invention, through the design of the vector field, the precision, smoothness and anti-interference capability of the ship in tracking the predetermined path under the complex sea condition are effectively improved, and the defects that the traditional method is poor in adaptability and is easily influenced by disturbance are overcome.

The invention provides a ship tracking multi-source information fusion method and a ship tracking multi-source information fusion system driven by variable long and short messages. The method comprises the following steps: acquiring ship observation data from at least one single-satellite information processing load; calculating a target position of the ship based on the ship observation data, and obtaining an estimated positioning error corresponding to the target position; determining a matched grid level from a preset global sea surface layered grid according to the estimated positioning error; according to the target position, determining a corresponding grid unit in the determined grid hierarchy, and obtaining a grid code of the grid unit; grid codes of the grid units serve as ship position information, and a position information message is generated and sent to a fusion processing terminal through a satellite short message communication link; and the fusion processing terminal performs multi-source information fusion on the grid codes from different single-satellite information processing loads or different time to generate a ship tracking trajectory.

The invention discloses a ship zero-sum game tracking control method based on digital-object dual-drive reinforcement learning. The method comprises the following steps: constructing a data physical dual-drive model; adopting a Lyapunov backstepping method to hierarchically design a ship track tracking feedforward controller; defining a performance index function and a Hamiltonian-Jacobian-Ansaxophone equation based on an error system and a zero-sum game theory, and designing an optimal feedback controller and a worst interference strategy; a fuzzy logic system is utilized to approach an HJI equation solution, and Nash equilibrium of an output estimation value of an optimal feedback controller and a worst interference strategy is obtained; a feedforward controller, an optimal feedback controller and a worst interference strategy are fused, and a final ideal ship controller capable of directly executing an actual control instruction is obtained through constraint processing, so that ship track tracking is realized. The method solves the problems that the existing method generally has the inherent defects of high dependence on high-quality data, weak physical interpretability, insufficient generalization ability to complex working conditions and the like, and is difficult to meet the safety and reliability requirements of ship navigation.

The application discloses a ship zero-sum game tracking control method based on data-physical dual driving and reinforcement learning, which comprises the following steps: constructing a data-physical dual driving model; adopting a Lyapunov backstepping method to design a ship track tracking feedforward controller in layers; defining a performance index function and a Hamilton-Jacobi-Isaacs equation based on an error system and a zero-sum game theory, and designing an optimal feedback controller and a worst disturbance strategy; using a fuzzy logic system to approximate a solution of the HJI equation, and obtaining a Nash equilibrium of estimated values of outputs of the optimal feedback controller and the worst disturbance strategy; and fusing the feedforward controller, the optimal feedback controller and the worst disturbance strategy, and performing constraint processing to obtain a final ideal ship controller which can directly execute an actual control instruction, so that ship track tracking is realized. The application solves the problems of inherent defects of existing methods, such as high dependence on high-quality data, weak physical interpretability, insufficient generalization ability for complex working conditions and the like, and cannot meet the safety and reliability requirements of ship navigation.

This invention provides an adaptive fuzzy tracking control method for unmanned surface vessels (USVs) with state quantization and input quantization. The method includes: acquiring information about the surrounding environment and sea state of other vessels; establishing a kinematic and dynamic model of the USV; designing a system control law based on an adaptive backstepping method; approximating uncertainties in the control system using a fuzzy logic system and designing an adaptive law to ensure the stability of the control system; quantizing the state variables and input variables in the control system using a uniform quantizer, and using the quantized state feedback information for the design of the USV trajectory tracking controller; proving the boundedness of the error between quantized and unquantized variables in the closed-loop control system using a recursive method; and proving the stability of the designed fuzzy adaptive feedback tracking control system with state quantization and input quantization, considering both state quantization and input quantization, based on Lyapunov stability theory.

The application relates to a multi-target association fusion method based on rules and a prediction model, and belongs to the field of target data association information fusion; prior knowledge rules based on a knowledge base are constructed; a set of data points required for model training is obtained by adopting a Douglas-Puke algorithm, the data amount is simplified while the geometric shape is reserved, and trajectory feature extraction is realized; a ship track prediction model based on a time convolution network (TCN) is constructed and is recorded as a TCN ship track prediction model; the extracted trajectory features are taken as inputs of the TCN ship track prediction model, and a TCN network prediction result is output; a target association method based on the prior knowledge rules and the TCN network prediction model is established, and targets are determined; the application realizes improvement of the accuracy of multi-target association fusion and ensures the real-time performance of the association fusion.

The invention belongs to the technical field of power generation, and particularly discloses a ship track power generation system which comprises a control system, a main ship body and a track ship body. The main ship body comprises a bottom supporting body and two sets of side supporting bodies, the rail ship body comprises a top sliding body and a warehouse, the two sides of the top sliding body can slide on the tops of the two sets of side supporting bodies respectively, and two sets of sliding assemblies are arranged at the bottom of the top sliding body; each sliding assembly comprises a plurality of idler wheels arranged in the length direction of the corresponding side face supporting body, and hub generators are arranged in the idler wheels. When the main ship body sails in the water, the warehouse is subjected to resistance, and the larger the resistance is, the larger the generating capacity is. According to the technical scheme, the main ship body and the rail ship body form a child-mother ship, the main ship body is propelled in water through water drained by the rear drainage channel, and no propeller is needed.

The invention relates to the technical field of ships, and provides a method and a system for identifying a forged ship trajectory, electronic equipment and a storage medium. The identification method for the ship counterfeit trajectory comprises the following steps: acquiring AIS data and a graph auto-encoder model; extracting a ship track of each ship and ship information of each ship from the AIS data; according to the ship trajectory and the ship information of each ship, constructing a cooperative behavior map among the plurality of ships; determining an abnormal value of each ship trajectory according to the cooperative behavior map and a map auto-encoder model; and determining a forged trajectory according to the abnormal value of each ship trajectory. According to the method, the cooperative behavior graph between the ships is constructed, and the graph auto-encoder model is utilized to automatically learn the cooperative rule of the normal ship group so as to identify the forged track, so that the method does not depend on manual labels any more, and the forged track can be automatically found under large-scale AIS data without labels.

The invention provides a ship track prediction method and system in a multi-ship interaction environment, and a storage medium. The method comprises the following steps: acquiring historical track data of a target ship and surrounding ships thereof, and preprocessing the historical track data; based on the preprocessed historical track data, constructing a dynamic directed graph sequence of the target ship under multiple time slices; the dynamic directed graph sequence is input into a graph neural network for feature aggregation, and feature representation of the target ship under each time slice is obtained; inputting the feature representation into a prediction network to obtain a track prediction result of the target ship; according to the method, the ship can effectively cope with the ship interaction problem in the dynamic evolution environment, result errors caused by neglecting the multi-ship interaction problem in the dynamic environment are avoided, and meanwhile, the method has higher robustness, accuracy and effectiveness compared with other schemes when facing scenes such as complex and high-density channels.