Ship trajectory state identification method and system, electronic device and storage medium

Abstract

The present application relates to the technical field of ship traffic management, and provides a ship trajectory state recognition method, system, electronic device and storage medium, the method comprising: acquiring a trajectory sequence of a ship in a continuous time period; matching a berthing level corresponding to each trajectory point; determining a trajectory direction value between adjacent two trajectory points based on the berthing levels of the adjacent two trajectory points, and generating a trajectory direction sequence; based on the trajectory direction sequence, segmenting the trajectory sequence into multiple trajectory sub-sequences with different trajectory directions; and performing ship trajectory state recognition on each trajectory sub-sequence.The ship trajectory state recognition method provided by the present application determines the direction change between trajectory points through the classification of the distance from the trajectory points to the coastline, further combines the trajectory direction change to segment the trajectory sequence, recognizes each segment, is especially suitable for the recognition of long-time sequence trajectories, effectively avoids the influence of mixed trajectory states of multiple segments, and can accurately recognize the ship trajectory state corresponding to each trajectory sub-sequence.

Description

Technical Field

[0001] The present invention relates to the technical field of ship traffic management, and in particular to a ship trajectory state recognition method, system, electronic equipment and storage medium. Background Art

[0002] With the booming shipping industry, demand for ships continues to rise, with ships gradually becoming more intelligent, larger, and more diverse. This has led to a continuous increase in the number of ships on shipping routes, especially in the waters around ports and nearshore waters. The increased load and the increasing risks faced by ships navigating nearshore waters have also increased. Therefore, identifying ship trajectory status has become particularly important. Accurate ship trajectory status identification not only helps ensure ship navigation safety and reduce the occurrence of maritime accidents, but also optimizes the utilization of waterway resources and improves shipping efficiency.

[0003] Identifying a ship's trajectory status requires analysis of its navigation data. Ships must be equipped with and have their Automatic Identification System (AIS) activated. This system automatically broadcasts spatiotemporal information, including the ship's name, Maritime Mobile Service Identity (MMSI), latitude and longitude, and navigation timestamp, at regular intervals. This information is collected and stored in a database, providing the lowest-cost and most comprehensive source of dynamic ship data. While ship-mounted AIS can provide a wealth of navigation data, effectively utilizing this data for accurate trajectory identification remains a challenge.

[0004] Existing ship trajectory state identification methods usually use traditional physics statistical models. Although the model is simplified, it can only basically meet the prediction scenarios of short-term ship trajectories. The prediction of long-term ship trajectories is not accurate and may be affected by the mixed effects of multiple trajectory states such as "docking first and then leaving the shore". Summary of the Invention

[0005] The present invention provides a method, system, electronic device and storage medium for identifying ship trajectory status, which are used to solve the problem of low accuracy in the prior art in predicting long time series of ship trajectory.

[0006] In a first aspect, the present invention provides a method for identifying a ship's trajectory state, comprising:

[0007] Acquire a trajectory sequence of the ship within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order;

[0008] Matching the corresponding landing level of each track point based on the distance from each track point to the coastline;

[0009] Based on the landing levels corresponding to the two adjacent track points, the track direction value between the two adjacent track points is determined to generate a track direction sequence;

[0010] Based on the trajectory direction sequence, dividing the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions;

[0011] Ship trajectory state recognition is performed on each of the trajectory subsequences to obtain a ship trajectory state recognition result corresponding to each of the trajectory subsequences of the ship.

[0012] In one embodiment, determining the track direction value between two adjacent track points based on the docking levels respectively corresponding to the two adjacent track points includes:

[0013] The difference between the landing grades corresponding to two adjacent track points is calculated to obtain the difference between the landing grades of the two adjacent track points;

[0014] Based on the difference in the landing level between two adjacent track points, the track direction value between the two adjacent track points is determined.

[0015] In one embodiment, determining the track direction value between two adjacent track points based on the difference in the landing level between the two adjacent track points includes:

[0016] If the difference in the landing level between two adjacent track points is greater than 0, the track direction value between the two adjacent track points is determined to be a first track direction value; the first track direction value is used to indicate that the track point state is an offshore state;

[0017] If the difference in the docking level between two adjacent track points is less than 0, the track direction value between the two adjacent track points is determined to be a second track direction value; the second track direction value is used to indicate that the track point state is a docking state;

[0018] If the difference in the docking level between two adjacent track points is equal to 0, the track direction value between the two adjacent track points is determined to be a third track direction value; the third track direction value is used to indicate that the track point state is no state change.

[0019] In one embodiment, generating a trajectory direction sequence includes:

[0020] generating a first intermediate processing sequence according to a trajectory direction value between two adjacent trajectory points;

[0021] forward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a second intermediate processing sequence;

[0022] filling backward each of the third track direction values in the first intermediate processing sequence to obtain a third intermediate processing sequence;

[0023] if the track direction values at any same position between the second intermediate processing sequence and the third intermediate processing sequence are different, resetting the track direction values as the third track direction values to obtain a track direction sequence.

[0024] In an embodiment, the splitting the track sequence into a plurality of track sub-sequences with different track directions based on the track direction sequence comprises:

[0025] traversing all the track direction values in the track direction sequence to determine positions where track direction changes occur;

[0026] splitting the track sequence into a plurality of track sub-sequences with different track directions by taking the positions where track direction changes occur as splitting points;

[0027] if the number of track points in any of the sub-sequences is less than a preset number, merging the sub-sequence into an adjacent sub-sequence to obtain a plurality of track sub-sequences with different track directions.

[0028] In an embodiment, when performing ship track state recognition on each of the track sub-sequences to obtain ship track state recognition results corresponding to each of the track sub-sequences of the ship, the following steps are performed for each track sub-sequence:

[0029] performing ship track state recognition on the track sub-sequence based on parameters associated with the berthing levels corresponding to the track points;

[0030] if the recognition is successful, generating the ship track state recognition results corresponding to the track sub-sequence of the ship;

[0031] if the recognition fails, performing ship track state recognition on the track sub-sequence based on parameters associated with the off-shore distances of the track points to generate the ship track state recognition results corresponding to the track sub-sequence of the ship.

[0032] In an embodiment, the parameters associated with the berthing levels corresponding to the track points include a first parameter, a second parameter, a third parameter, a fourth parameter, and a fifth parameter; the first parameter is a difference between the maximum berthing level and the minimum berthing level corresponding to the track points in the track sub-sequence; the second parameter is the berthing level corresponding to the starting track point in the track sub-sequence; the third parameter is the berthing level corresponding to the ending track point in the track sub-sequence; the fourth parameter is a difference between the third parameter and the second parameter; and the fifth parameter is the average speed of the ship in the track sub-sequence.

[0033] The parameters associated with the offshore distance of the trajectory point include the fifth parameter, the sixth parameter and the seventh parameter; the sixth parameter is the slope value of the curve formed by the offshore distance of the trajectory point in the trajectory subsequence changing with time; the seventh parameter is the difference between the offshore distance of the ending trajectory point and the offshore distance of the starting trajectory point in the trajectory subsequence.

[0034] In a second aspect, the present invention further provides a ship trajectory state identification system, comprising:

[0035] An acquisition module is used to acquire a trajectory sequence of a ship within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order;

[0036] A docking level matching module, configured to match the docking level corresponding to each track point based on the distance from each track point to the coastline;

[0037] A track direction determination module, configured to determine a track direction value between two adjacent track points based on the docking levels respectively corresponding to the two adjacent track points, and generate a track direction sequence;

[0038] a segmentation module, configured to segment the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions based on the trajectory direction sequence;

[0039] The ship track state recognition module is used to perform ship track state recognition on each of the track subsequences to obtain a ship track state recognition result corresponding to each of the track subsequences of the ship.

[0040] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the above-described methods for identifying ship trajectory status are implemented.

[0041] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of any of the above-mentioned ship trajectory status identification methods are implemented.

[0042] The ship trajectory state identification method, system, electronic device, and storage medium provided by the present invention classify the distance from the trajectory point to the coastline, can measure the relative positional relationship between the ship trajectory point and the coastline, determine the trajectory direction value between the trajectory points based on the docking level of the trajectory point, can display the navigation trend and direction change of the ship, facilitate more intuitive and accurate identification of the ship trajectory state, reduce the number of parameters for determining the ship trajectory state, and further combine the direction change between the trajectory points to segment the trajectory sequence, subdivide the overall trajectory of the ship in continuous time into multiple trajectory subsequences with different trajectory directions, and then perform separate ship trajectory state identification for each trajectory subsequence. This method is particularly suitable for identifying long-time trajectory sequences, effectively avoids the influence of mixed trajectory states of multiple segments, and can accurately identify the ship trajectory state corresponding to each trajectory subsequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 This is one of the flow charts of the ship trajectory state identification method provided by the present invention.

[0045] Figure 2 This is the second flow chart of the ship trajectory state identification method provided by the present invention.

[0046] Figure 3 It is a schematic diagram of the coastline and the shore-level ring zone provided by the present invention.

[0047] Figure 4 This is one of the trajectory diagrams of a ship in a docked or stationary state provided by the present invention.

[0048] Figure 5 This is the second schematic diagram of the trajectory of a ship in a docked or stationary state provided by the present invention.

[0049] Figure 6 This is the third schematic diagram of the trajectory of a ship in a docked or stationary state provided by the present invention.

[0050] Figure 7 One of the trajectory schematic diagrams of a ship docking state provided by the present invention.

[0051] Figure 8 The second schematic diagram of the trajectory of a ship docking provided by the present invention.

[0052] Figure 9One of the trajectory schematic diagrams of a ship in an offshore state provided by the present invention.

[0053] Figure 10 The second schematic diagram of the trajectory of a ship in an offshore state provided by the present invention.

[0054] Figure 11 One of the trajectory diagrams of various states of a ship provided by the present invention.

[0055] Figure 12 The second schematic diagram of the trajectory of a ship in various states provided by the present invention.

[0056] Figure 13 The third schematic diagram of the trajectory of a ship in various states provided by the present invention.

[0057] Figure 14 The fourth schematic diagram of the trajectory of a ship in various states provided by the present invention.

[0058] Figure 15 It is a structural diagram of the ship trajectory state identification system provided by the present invention.

[0059] Figure 16 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0060] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0061] The terms "first," "second," and the like in the present invention are used to distinguish similar objects and are not used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present invention can be implemented in orders other than those illustrated or described herein.

[0062] The following combination Figures 1-16 The present invention describes a method, system, electronic device and storage medium for identifying a ship's trajectory state.

[0063] The ship trajectory status identification method provided by the embodiment of the present invention is based on the three steps of "coast distance classification + ship trajectory direction sequence segmentation + two-stage status discrimination". Based on the real-time positioning data of the ship's AIS, by analyzing the dynamic information of the port area such as the distance between the ship and the coastline, speed changes, and the ship's berthing level, it automatically identifies the trajectory status of the ship, such as being away from the coastline, close to the coastline, and docked at the port. At the same time, according to the ship trajectory point data, multi-dimensional information of each trajectory point is attached, including detailed information such as ship speed, serial number of the nearest coastline point, distance from the nearest coastline point, latitude and longitude of the nearest coastline point, ship berthing level, and ship trajectory point status, providing detailed ship trajectory data for port scheduling, ship monitoring, and subsequent data analysis.

[0064] In addition, the ship trajectory state recognition method provided in the embodiment of the present invention is implemented based on the ship trajectory state recognition system. Therefore, the embodiment of the present invention uses the ship trajectory state recognition system as the execution body to specifically describe the ship trajectory state recognition method.

[0065] Combine Figure 1 and Figure 2 , Figure 1 This is one of the flow charts of the ship trajectory state identification method provided by the present invention. Figure 2 This is the second flow chart of the ship trajectory state identification method provided by the present invention.

[0066] like Figure 1 As shown, the ship trajectory state identification method includes the following steps:

[0067] Step 101: Obtain a trajectory sequence of a ship within a continuous time period.

[0068] Specifically, the ship trajectory data provided by AIS is typical spatiotemporal data, with spatiotemporal characteristics as its fundamental attributes. Spatial characteristics reflect the geographic environment and position distribution of the vessel's behavior, while temporal characteristics reflect the timing and periodicity of the behavior. Furthermore, ship trajectory data also contains rich information about the vessel's motion, such as position, speed, and heading, which comprehensively describes the motion characteristics of the vessel's behavior. Therefore, the ship trajectory data provided by AIS provides powerful data support for maritime traffic management and the automated and intelligent identification of vessel trajectory status.

[0069] If a vessel meets the weight and vessel type requirements, it must be equipped with and continuously powered on an AIS device. The system will acquire a large amount of vessel trajectory data from the AIS device. To avoid the influence of intersecting vessel trajectories and vessel trajectories at different times, the acquired vessel trajectory data is processed by vessel and time period before vessel trajectory status identification is performed.

[0070] For per-ship processing, the trajectory data corresponding to each ship is extracted from a large amount of ship trajectory data according to a fixed MMSI of each ship, and then the trajectory points in the extracted single-ship trajectory data are sequentially sorted according to the time of capture. The trajectory data processing and trajectory state recognition of all ships are the same, and therefore, the following content is described only for a single ship.

[0071] For time period processing, the extracted trajectory data of a single ship can be discontinuous, and therefore, the time period processing operation is only performed on discontinuous trajectories of the same ship to obtain trajectory data of the single ship in multiple continuous time periods. If the AIS continuously reports the trajectory data (i.e., the time difference between the trajectory points is less than the time threshold), no segmentation processing is performed. Specifically, the time difference t between each trajectory point and the previous trajectory point is calculated. If the time difference t is greater than or equal to the configured time threshold, it is considered to be a new trajectory segment. If the time difference t is less than the configured time threshold, no time period segmentation operation is performed. The foregoing process is repeated until each trajectory point is traversed. Each continuous motion segment of the single ship is separately labeled with a sequence number.

[0072] For example, the trajectory timestamp of a ship is 1:00-3:00, 5:00-7:00, and the missing 2 hours of reporting data is exactly equal to the time threshold. It is considered that the trajectory data of the ship is not continuously reported, and therefore, the trajectory data needs to be divided into two segments.

[0073] After all the AIS-reported ship trajectory data is processed per ship and per time period, the trajectory state of each continuous motion segment of the single ship is recognized, thereby effectively avoiding the influence of the cross-ship trajectory and the influence of the ship trajectory in different time periods.

[0074] Step 102, based on the distance of each trajectory point to the coastline, matching the corresponding berthing level of each trajectory point.

[0075] Specifically, by combining the AIS dynamic trajectory data and the port geographic information (coastline, etc.), the automatic recognition of the ship trajectory can be realized. Therefore, first, the corresponding coastline data of the actual island or port to be configured is drawn, which is connected by multiple coastline points, and then the coastline data is imported into the system, and each coastline point (including its latitude and longitude) is extracted and labeled for subsequent traversal calculation of the ship and state analysis.

[0076] To intuitively and accurately recognize the state of the ship trajectory, concentric buffer rings are constructed within a predetermined range based on the coastline, and are divided into multiple berthing levels according to a predetermined distance to reflect the degree of proximity of the trajectory point to the coastline.

[0077] Figure 3 This is a schematic diagram of the coastline and the shore-level ring zone provided by the present invention. Figure 3 As shown, optionally, based on a certain section of coastline, concentric buffer rings are constructed within the range of 0 to 2.5 km (kilometers), and divided into five shore-reaching levels at every 500 m (meters):

[0078] Landing level 1: 0~500m from the shore, closest to the shore, highest priority, marked in red;

[0079] Landing Level 2: 500-1000m offshore, lower priority than Landing Level 1, marked in orange;

[0080] Landing level 3: 1000-1500m offshore, lower priority than landing level 2, marked in yellow;

[0081] Landing Level 4: 1500-2000m offshore, lower priority than Landing Level 3, marked in grass green;

[0082] Landing level 5: 2000~2500m offshore, farthest from the shore, lowest priority, marked in green.

[0083] Furthermore, the minimum distance dist_m (unit: meter) between each track point in the track sequence and the coastline is calculated, and then the multidimensional space retrieval algorithm (K-Dimensional Tree, KD-Tree) is used to quickly extract the serial number and longitude and latitude information of the coast point closest to the track point, and save it to facilitate the visualization of subsequent ship output results.

[0084] According to the minimum distance dist_m from each track point to the coastline in the track sequence, dist_m is mapped to the preset landing level, thereby matching the landing level corresponding to each track point.

[0085] In addition, for the ship's trajectory sequence, it is also necessary to calculate the Euclidean distance between the i-th trajectory point and the i-1-th trajectory point. , divided by the time difference between two adjacent trajectory points , you can get the navigation speed of the i-th trajectory point , which is used for subsequent trajectory status analysis of the ship.

[0086] Step 103: determining the track direction value between the two adjacent track points based on the landing levels corresponding to the two adjacent track points, and generating a track direction sequence;

[0087] Step 104 : Based on the trajectory direction sequence, divide the trajectory sequence into a plurality of trajectory sub-sequences with different trajectory directions.

[0088] Specifically, the berthing grade is divided according to the distance from the ship's trajectory point to the coastline. Different berthing grades represent different offshore states of the ship. By comparing the berthing grades of two adjacent trajectory points, the change in the distance between the ship and the coastline at these two moments can be determined, providing basic information for judging the ship's sailing direction.

[0089] Therefore, based on the comparison of the docking levels corresponding to two adjacent track points, the track direction value between the two adjacent track points is determined, and different track direction values ​​correspond to different sailing directions. The track direction value specifically includes a first track direction value, a second track direction value, and a third track direction value. The first track direction value is used to indicate that the track point state is offshore, that is, the ship's sailing direction during this short period of time is away from the shore; the second track direction value is used to indicate that the track point state is docked, that is, the ship's sailing direction during this short period of time is towards the shore; the third track direction value is used to indicate that the track point state is unchanged, that is, the ship is not sailing during this short period of time.

[0090] By sequentially calculating the trajectory direction values ​​between two adjacent trajectory points in the trajectory sequence, a trajectory direction sequence can be generated. This trajectory direction sequence reflects the trajectory direction changes of the ship throughout the entire voyage process in a continuous time period, which helps in subsequent trajectory status analysis of the ship.

[0091] Since a continuous and complete ship trajectory may have multiple state changes such as "first docking, then leaving" or "first leaving, then docking", in order to better identify the ship's trajectory state reversal, the continuous and complete ship trajectory is first divided into sub-segments. After the sub-segmentation, the docking and leaving state is judged, and all state changes of the ship can be identified.

[0092] Traverse the trajectory direction sequence, determine the location where the trajectory direction changes, generate a set of flip points flip_points, and use these flip points as segmentation points to split the trajectory sequence into multiple trajectory subsequences with different trajectory directions. Each trajectory subsequence represents a navigation phase of the ship in a specific trajectory direction.

[0093] Step 105: Perform ship trajectory state recognition on each of the trajectory subsequences to obtain a ship trajectory state recognition result corresponding to each of the trajectory subsequences of the ship.

[0094] Specifically, to identify ship trajectory states, several state identification index parameters are defined, including parameters associated with the corresponding berthing level of the trajectory point and parameters associated with the distance from the shore. A rough judgment is first performed on each trajectory subsequence using the parameters associated with the berthing level. If the rough judgment succeeds, the ship trajectory state identification result corresponding to each trajectory subsequence is directly output. If the rough judgment fails, a detailed judgment is performed on each trajectory subsequence using the parameters associated with the distance from the shore, and the ship trajectory state identification result corresponding to each trajectory subsequence is then output.

[0095] By performing ship trajectory state recognition on each trajectory subsequence, the trajectory state of the ship in the navigation phase represented by each trajectory subsequence can be determined.

[0096] The ship trajectory state identification method provided by the present invention grades the distance from the trajectory point to the coastline, can measure the relative positional relationship between the ship trajectory point and the coastline, and determines the trajectory direction value between the trajectory points based on the docking level of the trajectory point. It can show the navigation trend and direction change of the ship, facilitate more intuitive and accurate identification of the ship trajectory state, reduce the number of parameters for judging the ship trajectory state, and further combine the direction change between the trajectory points to segment the trajectory sequence, subdivide the overall trajectory of the ship in continuous time into multiple trajectory subsequences with different trajectory directions, and then perform separate ship trajectory state identification for each trajectory subsequence. The method is particularly suitable for identifying long-time trajectory sequences, effectively avoids the influence of mixed trajectory states of multiple segments, and can accurately identify the ship trajectory state corresponding to each trajectory subsequence.

[0097] In some embodiments, based on step 103, determining the track direction value between two adjacent track points based on the docking levels corresponding to the two adjacent track points includes:

[0098] The difference between the landing grades corresponding to two adjacent track points is calculated to obtain the difference between the landing grades of the two adjacent track points;

[0099] Based on the difference in the landing level between two adjacent track points, the track direction value between the two adjacent track points is determined.

[0100] Specifically, after the entire time-sorted trajectory sequence, the landing level corresponding to each trajectory point is extracted to form a landing level sequence. The landing level sequence is smoothed to eliminate isolated noise jumps in the trajectory points.

[0101] Optionally, a 3-point median smoothing process is performed on each docking level in the docking level sequence. The principle of 3-point median smoothing is that for each docking level in the sequence, the "median" of the three values ​​of the docking level itself and its left and right neighboring levels is used to replace the current docking level, so as to achieve the purpose of removing noise and retaining the overall trend. If the first docking level or the last docking level lacks a neighboring level, the missing neighboring level will be filled with 0, and finally a smoothed and denoised docking level sequence Level_med can be obtained. The docking levels used in subsequent calculations of trajectory direction values, identification of ship trajectory status, etc. are also the docking levels in Level_med after smoothing and denoising.

[0102] Furthermore, the difference calculation is performed on each landing level in Level_med in chronological order to obtain the landing level difference between two adjacent trajectory points.

[0103] The track direction value between the two adjacent track points is determined based on the numerical value of the difference in the landing level between the two adjacent track points.

[0104] Specifically, if the difference in the docking level between two adjacent track points is greater than 0, the track direction value between the two adjacent track points is determined to be the first track direction value, recorded as direction=+1, indicating that the track point state is offshore.

[0105] If the difference in the docking level between two adjacent track points is less than 0, the track direction value between the two adjacent track points is determined to be the second track direction value, recorded as direction=-1, indicating that the track point state is the docking state.

[0106] If the difference in the landing level between two adjacent track points is 0, the track direction value between the two adjacent track points is determined to be the third track direction value, recorded as direction=0, indicating that the track point state is no state change.

[0107] The embodiment of the present invention determines the trajectory direction value based on the difference in the docking level between two adjacent trajectory points, which can intuitively and effectively reflect the state changes of the ship during navigation, facilitates more intuitive and accurate identification of the ship's trajectory state, and reduces the number of parameters for judging the ship's trajectory state.

[0108] In some embodiments, based on step 103, generating a trajectory direction sequence includes:

[0109] generating a first intermediate processing sequence according to a trajectory direction value between two adjacent trajectory points;

[0110] forward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a second intermediate processing sequence;

[0111] Backward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a third intermediate processing sequence;

[0112] If the trajectory direction value of any identical position between the second intermediate processing sequence and the third intermediate processing sequence is different, the trajectory direction value is reset to the third trajectory direction value to obtain a trajectory direction sequence.

[0113] Specifically, the first intermediate processing sequence Direction_seq can be generated by arranging the track direction values ​​between two adjacent tracks in sequence.

[0114] Forward fill is performed on each third track direction value (i.e., each 0 value) appearing in the first intermediate processing sequence, that is, each third track direction value (i.e., each 0 value) is replaced in sequence with the nearest non-zero track direction value (i.e., +1 value or -1 value) in front of it, to obtain the second intermediate processing sequence dir_fwd.

[0115] For example, Direction_seq=[-1,-1,0,-1,-1,0,0,0,0,0,1,0,1,0,0,1], after forward processing dir_fwd=[-1,-1,-1,-1,-1,-1,-1,-1,-1,1,1,1,1,1,1].

[0116] Backward filling is performed on each third track direction value (i.e., each 0 value) appearing in the first intermediate processing sequence, that is, each third track direction value (i.e., each 0 value) is replaced in sequence with the nearest non-zero track direction value (i.e., +1 value or -1 value) behind it to obtain the third intermediate processing sequence dir_bwd.

[0117] For example, Direction_seq=[-1,-1,0,-1,-1,0,0,0,0,0,1,0,1,0,0,1], after backward processing dir_bwd=[-1,-1,-1,-1,1,1,1,1,1,1,1,1,1,1].

[0118] Furthermore, a bitwise comparison is performed on the second intermediate processing sequence and the third intermediate processing sequence.

[0119] If the trajectory direction value of the second intermediate processing sequence and the third intermediate processing sequence is the same at a certain same position, the trajectory direction value is retained.

[0120] If at a certain same position, the trajectory direction values ​​of the second intermediate processing sequence and the third intermediate processing sequence are different, indicating that there is a conflict in the results obtained by forward filling and backward filling, the trajectory direction value of the position is reset to the third trajectory direction value (that is, set to 0) to retain the "opposite sign sandwich" 0 value.

[0121] The first intermediate processing sequence Direction_seq is updated to the track direction sequence Direction_med through data processing processes such as forward filling, backward filling, and bitwise comparison to retain the “different-sign sandwich” 0 value.

[0122] The embodiments of the present invention generate intermediate sequences through forward filling and backward filling and compare and verify them, effectively solving the problem of misjudgment of trajectory direction mutations that may be caused by a single filling direction. By cross-validating the bidirectional filling results, inconsistent trajectory direction values ​​are corrected, significantly improving the accuracy of the trajectory direction sequence and avoiding state switching points that may be masked by traditional unidirectional filling, thereby enhancing the reliability of subsequent trajectory subsequence segmentation and state recognition.

[0123] In some embodiments, based on step 104, dividing the trajectory sequence into a plurality of trajectory sub-sequences with different trajectory directions based on the trajectory direction sequence includes:

[0124] Traversing all trajectory direction values ​​in the trajectory direction sequence to determine a position where a trajectory direction change occurs;

[0125] Taking the position where the trajectory direction changes as a segmentation point, the trajectory sequence is segmented into a plurality of subsequences with different trajectory directions;

[0126] If the number of trajectory points in any subsequence is less than a preset number, the subsequence is merged into an adjacent subsequence to obtain a plurality of trajectory subsequences with different trajectory directions.

[0127] Specifically, all trajectory direction values ​​in the trajectory direction sequence are traversed to determine the position where the trajectory direction changes. A flip point set flip_points is generated. Once a direction sign flip is detected, the flip point is used as a segmentation point to segment a subsequence, and the subsequence number sub_id is incremented by 1. This ensures that consecutive trajectory points with the same direction are always included in the same subsequence. Only when the direction sign changes are the points switched to a new subsequence, providing a clear segmentation basis for subsequent trajectory state determination.

[0128] After dividing multiple subsequences with different trajectory directions, determine the number of trajectory points in each subsequence. The number of trajectory points in the subsequence is used to determine whether the subsequence is valid. The number of trajectory points in any valid subsequence must reach a preset number. The preset number can be set as min_sub=3.

[0129] If the number of trajectory points in a subsequence is less than the preset number, it means that the trajectory of the subsequence is too short and the segmentation is invalid. The subsequence needs to be merged into the adjacent subsequence, preferably into the left adjacent subsequence. If the subsequence is at the beginning of the sequence (no left neighbor), it is merged into the right adjacent subsequence.

[0130] After the merging is completed, multiple valid trajectory subsequences with different trajectory directions can be obtained, and all trajectory subsequences are renumbered continuously.

[0131] The embodiment of the present invention detects the position of trajectory direction changes to perform subsequence segmentation and introduces a minimum trajectory point number constraint to merge short subsequences. This effectively solves the problem of over-segmentation caused by noise interference caused by trajectory point position drift or short-term directional fluctuations. It not only ensures the effective segmentation of ship trajectory state changes, but also avoids invalid subsequences caused by data jitter. It significantly improves the robustness and rationality of trajectory subsequence segmentation and provides a clear segmentation basis for subsequent ship trajectory state identification.

[0132] In some embodiments, based on step 105, when performing ship trajectory state recognition on each of the trajectory subsequences to obtain the ship trajectory state recognition result corresponding to each of the trajectory subsequences of the ship, the following steps are performed for each trajectory subsequence:

[0133] performing ship trajectory state identification on the trajectory subsequence based on parameters associated with the docking level corresponding to the trajectory point;

[0134] If the recognition is successful, generating a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship;

[0135] If the recognition fails, ship trajectory state recognition is performed on the trajectory subsequence based on parameters associated with the offshore distance of the trajectory point, and a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship is generated.

[0136] Specifically, in order to more accurately realize the recognition of the ship trajectory status, a two-stage judgment logic is adopted, which is divided into coarse judgment and fine judgment. Coarse judgment is performed first and then fine judgment. If the coarse judgment has given an identification result, the fine judgment will not be entered.

[0137] For the rough judgment process, multiple parameters associated with the landing level corresponding to the trajectory points are defined, including but not limited to the first parameter Level_span, the second parameter LeL0, the third parameter LeLN, the fourth parameter Delta, and the fifth parameter Average_speed. The first parameter Level_span is the difference between the maximum landing level and the minimum landing level corresponding to the trajectory points in the trajectory subsequence. For example, if the maximum landing level is 3 and the minimum landing level is 2, the difference is 1. The second parameter LeL0 is the landing level corresponding to the starting trajectory point in the trajectory subsequence. The third parameter LeLN is the landing level corresponding to the ending trajectory point in the trajectory subsequence. The fourth parameter Delta is the difference between LeLN and LeL0, that is, the difference between the landing level corresponding to the ending trajectory point and the landing level corresponding to the starting trajectory point. The fifth parameter Average_speed is the average speed of the ship in the trajectory subsequence.

[0138] For the detailed judgment process, multiple parameters related to the offshore distance of the track points are defined, including but not limited to the fifth parameter Average_speed, the sixth parameter Slope and the seventh parameter D_dist. The sixth parameter Slope is the linear regression slope after the least squares fitting of the complete track subsequence "time-distance", that is, the slope value of the curve formed by the offshore distance of the track points in the track subsequence changing with time is used to judge the direction of the ship's track. For example, a one-variable linear regression is performed on the track data (t, d) using the least squares method to obtain the coefficient [Slope, b], which satisfies ; The seventh parameter D_dist is the difference between the offshore distance of the ending track point and the offshore distance of the starting track point in the track subsequence, in meters.

[0139] First, a rough judgment is made on the trajectory subsequence based on the parameters associated with the landing level corresponding to the trajectory point.

[0140] The judgment process of the rough judgment is as follows:

[0141] (1) If the maximum docking level corresponding to the trajectory point in the trajectory subsequence is equal to the docking level 1, and the first parameter Level_span is 0, indicating that the ship is always docked and has no dynamic level, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be the docking shoreline state;

[0142] (2) If the maximum docking level corresponding to the trajectory point in the trajectory subsequence is equal to docking level 2, and the first parameter Level_span is 0, and the fifth parameter Average_speed is less than the preset speed threshold Low_speed, such as Low_speed = 0.4m / s, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be stationary;

[0143] (3) If the fourth parameter Delta ≥ 1, that is, the ship trajectory crosses a shore level and the shore level corresponding to the end trajectory point is greater than the shore level corresponding to the start trajectory point, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be the leaving coastline state;

[0144] (4) If the fourth parameter Delta ≤ -1, that is, the ship trajectory crosses a shore level and the shore level corresponding to the end trajectory point is less than the shore level corresponding to the start trajectory point, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be close to the coastline state.

[0145] If the ship trajectory subsequence does not determine the ship trajectory status recognition result after the above four rough judgments, it enters the detailed judgment; if the ship trajectory subsequence has determined the ship trajectory status recognition result after the above four rough judgments, it jumps out of the judgment.

[0146] The judgment process is as follows:

[0147] (1) If the fifth parameter Average_speed in the trajectory subsequence is less than the preset speed threshold Low_speed, the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be stationary;

[0148] (2) If the sixth parameter Slope in the trajectory subsequence is greater than the first preset slope threshold, such as Slope>0.015, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be the state of leaving the coastline;

[0149] (3) If the sixth parameter Slope in the trajectory subsequence is less than the second preset slope threshold, such as Slope<-0.015, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be close to the coastline state;

[0150] (4) If the sixth parameter Slope in the trajectory subsequence does not change significantly, that is, the slope approaches 0, and the docking status cannot be determined, the seventh parameter D_dist is used for a fallback judgment, and the distance threshold can be set to DIST_THR = 100 meters;

[0151] If the seventh parameter D_dist in the trajectory subsequence is greater than DIST_THR, the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be the state of leaving the coastline;

[0152] If the seventh parameter D_dist in the trajectory subsequence is less than -DIST_THR, then the ship trajectory state recognition result corresponding to the trajectory subsequence is determined to be close to the coastline state;

[0153] If the ship trajectory subsequence does not produce a ship trajectory state identification result after the above four detailed judgments, the ship trajectory state identification result corresponding to the trajectory subsequence is determined to be other states.

[0154] The embodiment of the present invention uses a two-stage decision-making and complementary ship trajectory state recognition method to effectively avoid the recognition blind spot problem of a single feature in complex scenarios. It covers a comprehensive range of ship trajectory states and can accurately identify multiple ship trajectory states, facilitating ship monitoring in ports. In addition, the two-stage decision-making and judgment based on the segmented processing of the trajectory sequence not only greatly reduces the amount of ship trajectory data calculation and the complexity of judging the ship state, but also reduces the probability of missed and misjudgment of trajectory points.

[0155] It should be noted that in addition to the ship trajectory state recognition method based on traditional physics statistical models, the existing ship trajectory state recognition also adopts the ship trajectory state recognition method based on density clustering and the ship trajectory state recognition method based on deep learning.

[0156] Although the ship trajectory state identification method based on the traditional physics statistical model is simplified, it can only basically meet the prediction scenarios of short-term ship trajectories. It is inaccurate for the prediction of long-term ship trajectories. Moreover, when the time interval between trajectory points is too long, the uncertainty of this estimation model increases greatly.

[0157] The ship trajectory state recognition method based on density clustering relies too much on density clustering parameters, such as spatial threshold, time threshold, number of neighbor thresholds, etc. The clustering threshold parameters are sensitive and have poor portability. These threshold parameters need to change according to changes in ship scenarios, which consumes a lot of resources.

[0158] Although the deep learning-based ship trajectory state recognition method can significantly improve the accuracy of ship trajectory prediction and the ability to express complex interactions, it is highly dependent on large-scale, high-quality ship trajectory data. Training requires millions of continuous, synchronized AIS trajectories containing environmental characteristics. The model is prone to failure when data is missing, there is high noise, or samples in small ports are sparse. In addition, the algorithm results exhibit black box characteristics, cannot meet the requirements of interpretability, and have high hardware requirements and complexity.

[0159] Based on the above complete ship trajectory state recognition process, the defects of the existing technology can be effectively solved.

[0160] First, the ship trajectories of different time sequences and different states can be accurately segmented, avoiding the influence of multiple track mixtures such as "first docking and then leaving the shore" in the ship trajectory. The corresponding ship trajectory state can be accurately identified, reducing the dependence on the continuity of the ship trajectory time sequence, and solving the problem of ship trajectory misjudgment caused by AIS track point position drift and loss of report.

[0161] Secondly, the algorithm has strong robustness, low complexity, and small amount of trajectory point data demand. Through the division of the off-shore distance level buffer ring, the whole ship trajectory can be judged to give the whole trajectory state, and the discrete and few ship trajectory points can also be judged in detail. The two-stage ship trajectory prediction (first rough judgment and then detailed judgment) not only greatly reduces the calculation amount of ship trajectory data and the complexity of judging the ship state, but also reduces the missing and misjudgment probability of AIS trajectory points.

[0162] Furthermore, the algorithm has strong interpretability, wide coverage of trajectory state, and clear physical meaning of distance threshold, speed threshold, and sequence trend threshold, which is convenient for monitoring and auditing, and can give multiple trajectory states such as berthing, static, off-shore, and berthing.

[0163] The algorithm simulation results of the ship trajectory state recognition process are as follows Figure 4-Figure 14 .

[0164] Figure 4 is one of the ship berthing or static state trajectory schematic diagrams provided by the application; Figure 5 is the second ship berthing or static state trajectory schematic diagram provided by the application; Figure 6 is the third ship berthing or static state trajectory schematic diagram provided by the application.

[0165] Figure 7 is one of the ship berthing state trajectory schematic diagrams provided by the application; Figure 8 is the second ship berthing state trajectory schematic diagram provided by the application.

[0166] Figure 9 is one of the ship off-shore state trajectory schematic diagrams provided by the application; Figure 10 is the second ship off-shore state trajectory schematic diagram provided by the application.

[0167] Figure 11 is one of the ship multiple state trajectory schematic diagrams provided by the application; Figure 12 is the second ship multiple state trajectory schematic diagram provided by the application; Figure 13 is the third ship multiple state trajectory schematic diagram provided by the application; Figure 14 is the fourth ship multiple state trajectory schematic diagram provided by the application.

[0168] In order to better show the simulation results of the system algorithm, the ship historical trajectory is distinguished by different colors, and the state event types (off-shore, berthing, berthing, and static) are marked in the legend. And the berthing level of the ship is also marked by different colors, which more intuitively shows the relationship between the ship and the coastline.

[0169] The structure of the ship trajectory state recognition system provided by the present invention is described below. The ship trajectory state recognition system described below and the ship trajectory state recognition method described above can be referenced to each other.

[0170] Reference Figure 15 , Figure 15 It is a structural diagram of the ship trajectory state identification system provided by the present invention.

[0171] like Figure 15 As shown, the ship trajectory state identification system includes:

[0172] An acquisition module 1510 is configured to acquire a trajectory sequence of a vessel within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order;

[0173] The docking level matching module 1520 is used to match the docking level corresponding to each track point based on the distance from each track point to the coastline;

[0174] A track direction determination module 1530 is configured to determine a track direction value between two adjacent track points based on the landing levels corresponding to the two adjacent track points, and generate a track direction sequence;

[0175] a segmentation module 1540 for segmenting the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions based on the trajectory direction sequence;

[0176] The ship track state recognition module 1550 is configured to perform ship track state recognition on each of the track subsequences to obtain a ship track state recognition result corresponding to each of the track subsequences of the ship.

[0177] The ship trajectory state identification system provided by the present invention grades the distance from the trajectory point to the coastline, can measure the relative positional relationship between the ship trajectory point and the coastline, and determines the trajectory direction value between the trajectory points based on the docking level of the trajectory point. It can display the navigation trend and direction change of the ship, facilitate more intuitive and accurate identification of the ship trajectory state, reduce the number of parameters for judging the ship trajectory state, and further combine the direction change between the trajectory points to segment the trajectory sequence, subdivide the overall trajectory of the ship in continuous time into multiple trajectory subsequences with different trajectory directions, and then perform separate ship trajectory state identification for each trajectory subsequence. The system is particularly suitable for identifying long-time trajectory sequences, effectively avoiding the influence of mixed trajectory states of multiple segments, and can accurately identify the ship trajectory state corresponding to each trajectory subsequence.

[0178] Furthermore, the trajectory direction determination module 1530 is further configured to:

[0179] The difference between the landing grades corresponding to two adjacent track points is calculated to obtain the difference between the landing grades of the two adjacent track points;

[0180] Based on the difference in the landing level between two adjacent track points, the track direction value between the two adjacent track points is determined.

[0181] Furthermore, the trajectory direction determination module 1530 is further configured to:

[0182] If the difference in the landing level between two adjacent track points is greater than 0, the track direction value between the two adjacent track points is determined to be a first track direction value; the first track direction value is used to indicate that the track point state is an offshore state;

[0183] If the difference in the docking level between two adjacent track points is less than 0, the track direction value between the two adjacent track points is determined to be a second track direction value; the second track direction value is used to indicate that the track point state is a docking state;

[0184] If the difference in the docking level between two adjacent track points is equal to 0, the track direction value between the two adjacent track points is determined to be a third track direction value; the third track direction value is used to indicate that the track point state is no state change.

[0185] Furthermore, the trajectory direction determination module 1530 is further configured to:

[0186] generating a first intermediate processing sequence according to a trajectory direction value between two adjacent trajectory points;

[0187] forward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a second intermediate processing sequence;

[0188] Backward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a third intermediate processing sequence;

[0189] If the trajectory direction value of any identical position between the second intermediate processing sequence and the third intermediate processing sequence is different, the trajectory direction value is reset to the third trajectory direction value to obtain a trajectory direction sequence.

[0190] Furthermore, the segmentation module 1540 is further configured to:

[0191] Traversing all trajectory direction values ​​in the trajectory direction sequence to determine a position where a trajectory direction change occurs;

[0192] Taking the position where the trajectory direction changes as a segmentation point, the trajectory sequence is segmented into a plurality of subsequences with different trajectory directions;

[0193] If the number of trajectory points in any subsequence is less than a preset number, the subsequence is merged into an adjacent subsequence to obtain a plurality of trajectory subsequences with different trajectory directions.

[0194] Furthermore, the ship trajectory state identification module 1550 is further configured to:

[0195] performing ship trajectory state identification on the trajectory subsequence based on parameters associated with the docking level corresponding to the trajectory point;

[0196] If the recognition is successful, generating a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship;

[0197] If the recognition fails, ship trajectory state recognition is performed on the trajectory subsequence based on parameters associated with the offshore distance of the trajectory point, and a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship is generated.

[0198] It should be noted that the ship trajectory state identification system provided by the present invention can execute the ship trajectory state identification method described in any of the above embodiments during specific operation, which will not be described in detail in this embodiment.

[0199] Figure 16 Schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 16 As shown, the electronic device may include: a processor 1610, a communications interface 1620, a memory 1630, and a communication bus 1640. The processor 1610, the communications interface 1620, and the memory 1630 communicate with each other via the communication bus 1640. The processor 1610 may invoke logic instructions in the memory 1630 to execute a method for identifying a ship's trajectory state. The method includes: obtaining a trajectory sequence of a ship within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order; matching the corresponding shorebound level of each trajectory point based on the distance from each trajectory point to the coastline; determining the trajectory direction value between two adjacent trajectory points based on the shorebound levels corresponding to each of the two adjacent trajectory points to generate a trajectory direction sequence; segmenting the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions based on the trajectory direction sequence; and performing ship trajectory state identification on each of the trajectory subsequences to obtain a ship trajectory state identification result corresponding to each of the trajectory subsequences of the ship.

[0200] Furthermore, the logic instructions in the aforementioned memory 1630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0201] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the ship trajectory state identification method provided by the above-mentioned embodiments, the method including: obtaining a trajectory sequence of a ship in a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order; based on the distance of each of the trajectory points to the coastline, matching the corresponding berthing level of each of the trajectory points; based on the berthing levels corresponding to two adjacent trajectory points, determining the trajectory direction value between the two adjacent trajectory points, and generating a trajectory direction sequence; based on the trajectory direction sequence, dividing the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions; performing ship trajectory state identification on each of the trajectory subsequences, and obtaining the ship trajectory state identification results corresponding to each of the trajectory subsequences of the ship.

[0202] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the ship trajectory state identification method provided by the above-mentioned embodiments, the method comprising: obtaining a trajectory sequence of a ship within a continuous time period; the trajectory sequence comprises a plurality of trajectory points sorted in chronological order; based on the distance of each of the trajectory points to the coastline, matching the corresponding berthing level of each of the trajectory points; based on the berthing levels corresponding to two adjacent trajectory points, determining the trajectory direction value between the two adjacent trajectory points, and generating a trajectory direction sequence; based on the trajectory direction sequence, dividing the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions; performing ship trajectory state identification on each of the trajectory subsequences, and obtaining a ship trajectory state identification result corresponding to each of the trajectory subsequences of the ship.

[0203] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Those skilled in the art will be able to understand and implement the present invention without inventive effort.

[0204] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0205] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for identifying ship trajectory status, characterized in that: The ship trajectory state identification method comprises: Acquire a trajectory sequence of the ship within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order; Matching the corresponding landing level of each track point based on the distance from each track point to the coastline; Based on the landing levels corresponding to the two adjacent track points, the track direction value between the two adjacent track points is determined to generate a track direction sequence; Based on the trajectory direction sequence, dividing the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions; performing ship trajectory state recognition on each of the trajectory subsequences to obtain a ship trajectory state recognition result corresponding to each of the trajectory subsequences of the ship; The determining of the track direction value between the two adjacent track points based on the docking levels respectively corresponding to the two adjacent track points includes: The difference between the landing grades corresponding to two adjacent track points is calculated to obtain the difference between the landing grades of the two adjacent track points; Based on the difference in the landing level between two adjacent track points, the track direction value between the two adjacent track points is determined.

2. The ship trajectory state recognition method according to claim 1, characterized in that: The determining of the track direction value between the two adjacent track points based on the difference in the landing level between the two adjacent track points includes: If the difference in the landing level between two adjacent track points is greater than 0, the track direction value between the two adjacent track points is determined to be a first track direction value; the first track direction value is used to indicate that the track point state is an offshore state; If the difference in the docking level between two adjacent track points is less than 0, the track direction value between the two adjacent track points is determined to be a second track direction value; the second track direction value is used to indicate that the track point state is a docking state; If the difference in the docking level between two adjacent track points is equal to 0, the track direction value between the two adjacent track points is determined to be a third track direction value; the third track direction value is used to indicate that the track point state is no state change.

3. The ship trajectory state recognition method according to claim 2, characterized in that: Generating a trajectory direction sequence includes: generating a first intermediate processing sequence according to a trajectory direction value between two adjacent trajectory points; forward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a second intermediate processing sequence; Backward filling each of the third trajectory direction values ​​appearing in the first intermediate processing sequence to obtain a third intermediate processing sequence; If the trajectory direction value of any identical position between the second intermediate processing sequence and the third intermediate processing sequence is different, the trajectory direction value is reset to the third trajectory direction value to obtain a trajectory direction sequence.

4. The ship trajectory state recognition method according to claim 3, characterized in that: The step of dividing the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions based on the trajectory direction sequence includes: Traversing all trajectory direction values ​​in the trajectory direction sequence to determine a position where a trajectory direction change occurs; Taking the position where the trajectory direction changes as a segmentation point, the trajectory sequence is segmented into a plurality of subsequences with different trajectory directions; If the number of trajectory points in any subsequence is less than a preset number, the subsequence is merged into an adjacent subsequence to obtain a plurality of trajectory subsequences with different trajectory directions.

5. The ship trajectory state recognition method according to any one of claims 1 to 4, characterized in that: When performing ship trajectory state recognition on each of the trajectory subsequences to obtain the ship trajectory state recognition results corresponding to each of the trajectory subsequences of the ship, the following steps are performed for each trajectory subsequence: performing ship trajectory state identification on the trajectory subsequence based on parameters associated with the docking level corresponding to the trajectory point; If the recognition is successful, generating a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship; If the recognition fails, ship trajectory state recognition is performed on the trajectory subsequence based on parameters associated with the offshore distance of the trajectory point, and a ship trajectory state recognition result corresponding to the trajectory subsequence of the ship is generated.

6. The ship trajectory state recognition method according to claim 5, characterized in that: The parameters associated with the landing level corresponding to the track point include a first parameter, a second parameter, a third parameter, a fourth parameter, and a fifth parameter; the first parameter is the difference between the maximum landing level and the minimum landing level corresponding to the track points in the track subsequence; The second parameter is the docking level corresponding to the starting track point in the track subsequence; the third parameter is the docking level corresponding to the ending track point in the track subsequence; the fourth parameter is the difference between the third parameter and the second parameter; the fifth parameter is the average speed of the ship in the track subsequence; The parameters associated with the offshore distance of the trajectory point include the fifth parameter, the sixth parameter and the seventh parameter; the sixth parameter is the slope value of the curve formed by the offshore distance of the trajectory point in the trajectory subsequence changing with time; the seventh parameter is the difference between the offshore distance of the ending trajectory point and the offshore distance of the starting trajectory point in the trajectory subsequence.

7. A ship trajectory status recognition system, characterized in that: include: An acquisition module is used to acquire a trajectory sequence of a ship within a continuous time period; the trajectory sequence includes a plurality of trajectory points sorted in chronological order; A docking level matching module, configured to match the docking level corresponding to each track point based on the distance from each track point to the coastline; A track direction determination module, configured to determine a track direction value between two adjacent track points based on the docking levels respectively corresponding to the two adjacent track points, and generate a track direction sequence; a segmentation module, configured to segment the trajectory sequence into a plurality of trajectory subsequences with different trajectory directions based on the trajectory direction sequence; a ship trajectory state identification module, configured to perform ship trajectory state identification on each of the trajectory subsequences, and obtain a ship trajectory state identification result corresponding to each of the trajectory subsequences of the ship; The determining of the track direction value between the two adjacent track points based on the docking levels respectively corresponding to the two adjacent track points includes: The difference between the landing grades corresponding to two adjacent track points is calculated to obtain the difference between the landing grades of the two adjacent track points; Based on the difference in the landing level between two adjacent track points, the track direction value between the two adjacent track points is determined.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the ship trajectory status identification method according to any one of claims 1 to 6 are implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the ship trajectory status identification method according to any one of claims 1 to 6 are implemented.

Images