A mass ship trajectory adaptive reconstruction method considering ship motion state

Abstract

The application provides a mass ship trajectory adaptive reconstruction method considering ship motion state, comprising the following steps: S1, merging dynamic data and static data in a ship automatic identification system data into a trajectory data table; S2, dividing mass trajectory points into different voyage trajectory segments according to adjacent trajectory point time intervals based on a sliding window algorithm; S3, identifying and removing position, heading and speed outliers in each voyage trajectory segment; S4, identifying trajectory points in the voyage trajectory segment processed in S3 as turning motion state and straight sailing motion state based on a sliding window algorithm; S5, adaptively taking a specific linear interpolation method or a cubic spline interpolation method to reconstruct a trajectory segment according to the motion state characteristics of each motion segment, and merging the reconstructed each motion segment into a complete voyage trajectory segment. The application proposes a new idea of mass ship trajectory adaptive reconstruction, and can provide support for ship behavior pattern mining and marine traffic flow feature extraction.

Description

Technical Field

[0001] This invention relates to the field of maritime traffic safety technology, and more particularly to an adaptive reconstruction method for massive ship trajectories that takes into account the ship's motion state. Background Technology

[0002] Maritime trade plays a vital role in the process of trade globalization. With the increasing demand for maritime trade, ship traffic will grow accordingly, making maritime navigation safety management extremely challenging. Automatic Identification Systems (AIS), as sensors for maritime traffic monitoring, are widely used in maritime services and have accumulated massive amounts of dynamic and static information about ships. Mining and analyzing this massive ship navigation trajectory data can provide technical support for port planning and management, including ship navigation aids, identification of abnormal ship behavior, and maritime risk assessment. However, in practical applications, due to issues with positioning equipment, software design, and human operation, AIS data contains missing and outlier values. These outliers can affect maritime traffic monitoring, leading to misjudgments of the maritime traffic situation, and also hindering research on ship traffic pattern extraction, ship traffic behavior analysis, and traffic safety situation calculation based on AIS data.

[0003] Currently, ship trajectory reconstruction work focuses on designing methods for single trajectories, primarily encompassing two parts: trajectory outlier identification and trajectory repair. Outlier identification employs distance-based outlier detection and threshold detection based on human experience. These methods lack consideration for the ship's inherent characteristics, resulting in low adaptability of the discrimination criteria. Trajectory repair utilizes methods such as spline interpolation, linear regression, particle filtering, Kalman filtering, neural networks, and support vector machines. However, these methods lack consideration for the ship's motion state, applying a single repair method to trajectories under different motion states. Summary of the Invention

[0004] To address the aforementioned issues of existing ship trajectory reconstruction methods lacking consideration of ship attributes and motion states, and lacking the technical capacity to handle massive ship trajectory reconstruction, this invention provides an adaptive reconstruction method for massive ship trajectories that considers ship motion states. The massive trajectories are divided into different voyages based on the ship's MMSI and dwell time thresholds. Then, while considering the ship's attributes and motion states, adaptive reconstruction of ship trajectories is achieved using linear interpolation or cubic spline interpolation.

[0005] The technical means employed in this invention are as follows:

[0006] An adaptive reconstruction method for massive ship trajectories considering ship motion states includes the following steps:

[0007] S1. Merge the dynamic and static data in the Automatic Identification System (AIS) data into a trajectory data table, and sort the information in the trajectory data table in ascending order using the MMSI and the record time attribute column.

[0008] S2. Based on the ship's speed value, delete the ship's stop points in the trajectory data table after ascending sorting. Based on the sliding window algorithm, divide the points in the trajectory data table whose time interval is greater than the time threshold into different voyage trajectory segments.

[0009] S3. Identify and remove outliers in position, heading, and speed in each voyage trajectory segment;

[0010] S4. Based on the sliding window algorithm, the trajectory points in the voyage trajectory segment processed by S3 are divided into turning motion state and straight motion state. Adjacent trajectory points in the same motion state are merged into a straight motion segment or a turning motion segment.

[0011] S5. Based on the sliding window algorithm, different motion segments are identified. Based on the motion state characteristics of each motion segment, specific linear interpolation methods or cubic spline interpolation methods are adaptively adopted to reconstruct the trajectory segment. Each reconstructed motion segment is merged into a complete flight trajectory segment.

[0012] Furthermore, S1 specifically includes the following steps:

[0013] S11. Merging dynamic and static data information:

[0014] After decoding the Automatic Identification System (AIS) data, dynamic and static data information tables are obtained. The static data table is denoted as A, and the dynamic data table is denoted as B. Using the MMSI attribute columns as indexes, the attribute columns in table A are matched to table B and merged to form a new trajectory data table C. The attribute columns related to ship motion are retained.

[0015] S12. Trajectory data sorting:

[0016] The trajectory data table C is sorted according to the MMSI and recording time attribute columns to obtain spatiotemporal motion data reflecting the continuous movement of the ship.

[0017] Furthermore, in S11, the attributes related to ship motion specifically include MMSI, recording time, longitude, latitude, heading, speed, and ship length information.

[0018] Furthermore, S2 specifically includes the following steps:

[0019] Based on the time interval between adjacent trajectory points, the trajectory is divided into different voyage trajectory segments. First, the data rows in the ship trajectory dataset with a speed of 0 are deleted. Then, the trajectory of the same ship is extracted based on the MMSI. The time difference between adjacent trajectory points is calculated using the sliding window algorithm. Based on the dwell time threshold Δt, it is determined whether the two points belong to the same voyage. The voyage division operation in the previous step is applied to all MMSIs to achieve the division of all voyages in the trajectory dataset.

[0020] Furthermore, S3 specifically includes the following steps:

[0021] S31, Location noise identification:

[0022] For each voyage trajectory data T i,j Two adjacent points P n and P n+1 The distance Δd between n,n+1 It is obtained from the following formula:

[0023] Δd n,n+1 =ED(P n ,P n+1 )

[0024] Where ED(·,·) represents the Euclidean distance between two points and the average velocity between two adjacent points. It is obtained from the following formula:

[0025]

[0026] Where, Δd n,n+1 This represents the time interval between two adjacent points, if the average velocity between two adjacent points... Much greater than the highest speed during this voyage These two points are then determined to be location noise points, and P is set to... n and P n+1 Two deletions;

[0027] S32, Speed ​​Noise Recognition:

[0028] SOG at this point n Compare the velocities of two points before and after the trajectory point (SOG) n+1 and SOG n-1 If SOG n If the following formula is satisfied, the speed value can be determined to be abnormal, and the speed value can be deleted.

[0029]

[0030] Among them, v max This represents the maximum value described in the ITU for ship speed;

[0031] S33, Heading noise identification:

[0032] If the angular velocity of the trajectory point relative to the previous point is greater than θ max If the following formula is satisfied, then the trajectory point is an abnormal turning point and the speed value is deleted.

[0033]

[0034] Where, θ max Let v be the maximum turning angle, v be the ship speed, L be the ship length, k ∈ [2, 4], and P be the ship length. i Let P be the initial trajectory point. i+1 This is the final trajectory point.

[0035] Furthermore, S4 specifically includes the following steps:

[0036] The ship's motion state is divided into straight-line motion state and turning motion state. The straight-line motion state means that the ship is stable in a certain heading and undergoes uniform speed change motion. The turning motion state means that the ship achieves continuous change of direction by changing the steering gear angle. The rate of change of heading between two adjacent points is calculated using a sliding window to determine whether the ship is currently in a turning or straight-line state. Then, adjacent trajectory points in the same motion state are merged into a straight-line motion segment or a turning motion segment.

[0037] For each voyage trajectory data T i,j Two adjacent points P n and P n+1 The heading difference ΔCOG between n,n+1 It can be obtained from the following formula:

[0038] ΔCOG n,n+1 =|COG n+1 -COG n |

[0039] Among them, COG n+1 P represents n+1 The heading value, COG n P represents n The heading value, P n Rate of change of heading at point Δrot n It can be obtained from the following formula:

[0040]

[0041] When Δrot n Greater than the heading change rate threshold θ rot Then, it can be determined that P n The point represents the turning motion state. Otherwise, it is in a straight-line motion state.

[0042] Motion state of each point in the voyage trajectory Where n = 1, 2, ... k, it means that there are a total of k trajectory points for this voyage, and m = {s, t}, where s represents the straight-line state and t represents the turning state;

[0043] When the sliding window algorithm is used to identify adjacent trajectory points as belonging to the same motion state, the adjacent trajectory points are merged into the same motion trajectory segment. This allows the trajectory of a flight to be divided into motion segments under different motion states. Where x represents the number of movement segments that the voyage can be divided into, and y represents whether the movement segment is a straight movement segment or a turning movement segment.

[0044] Furthermore, S5 specifically includes the following steps:

[0045] The sliding window algorithm is used to count the number of trajectory points contained in different motion segments within each voyage trajectory. If the number of trajectory points in a motion segment is less than 3, the motion segment is merged with the previous motion segment to finally obtain different motion segments with obvious motion state characteristics.

[0046] For flight trajectories containing different motion segments, specific interpolation methods are selected to reconstruct the motion characteristics. For these trajectories, a sliding window algorithm is used to identify and obtain complete motion segments, and an interpolation method is adaptively selected to reconstruct the segments based on their motion state characteristics. For straight-line motion segments, a linear interpolation method is used to reconstruct the trajectory, as shown in the following formula:

[0047]

[0048] For the turning motion segment, the trajectory is reconstructed using cubic spline interpolation, as shown in the following formula; finally, all reconstructed motion segments within this voyage are merged to form a complete trajectory;

[0049]

[0050] Where t is the reconstruction time. and These are the coefficients of the linear interpolation function. and These are the coefficients of the cubic spline interpolation function.

[0051] The present invention also provides a storage medium comprising a stored program, wherein, when the program is executed, it performs any of the above-described adaptive reconstruction methods for massive ship trajectories that take into account the ship's motion state.

[0052] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes any of the above-described adaptive reconstruction methods for massive ship trajectories that take into account the ship's motion state through the computer program.

[0053] Compared with the prior art, the present invention has the following advantages:

[0054] Based on the actual process of ships completing transportation tasks, this invention adopts the "divide and conquer" approach to divide massive trajectories into different voyage trajectory segments according to MMSI and dwell time intervals, thereby improving the efficiency of ship trajectory data management, querying, and indexing.

[0055] This invention combines the ship's own attributes and motion characteristics to construct a method for identifying outliers in the ship's trajectory, which eliminates the interference of outliers on trajectory reconstruction and ensures the accuracy of the reconstructed trajectory.

[0056] This invention designs a motion trajectory segment division method that takes into account the motion state of a ship, which can divide the voyage trajectory into straight and turning motion trajectory segments, and adaptively use a specific interpolation method to repair the trajectory segments. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a flowchart of the method of the present invention.

[0059] Figure 2 This is a schematic diagram illustrating the identification of abnormal points in the ship trajectory according to the present invention. Figure 2 'a' is a scatter plot of trajectory anomalies. Figure 2 b is a curve of trajectory anomalies.

[0060] Figure 3 This is a schematic diagram illustrating the identification of abnormal ship speeds according to the present invention. Figure 3 'a' is a scatter plot of velocity anomalies. Figure 3 b is a curve of velocity anomalies.

[0061] Figure 4 This is a schematic diagram illustrating the identification of abnormal points in a ship's course according to the present invention. Figure 4 'a' is a scatter plot of heading anomalies. Figure 4 b is a curve diagram of the heading anomaly point.

[0062] Figure 5This is a schematic diagram of the ship trajectory point reconstruction results of the present invention.

[0063] Figure 6 This is a schematic diagram of the ship speed point reconstruction results of the present invention.

[0064] Figure 7 This is a schematic diagram of the ship's course point reconstruction result according to the present invention.

[0065] Figure 8 This is a schematic diagram of the original trajectories of a large number of ships in this invention.

[0066] Figure 9 This is a schematic diagram of the massive ship trajectory reconstruction results of the present invention. Detailed Implementation

[0067] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0068] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0069] like Figure 1 As shown, this invention provides an adaptive reconstruction method for massive ship trajectories that considers the ship's motion state, including the following steps:

[0070] S1. Merge the dynamic and static data from the Automatic Identification System (AIS) data, and sort the information in the trajectory data table in ascending order using MMSI and recording time. Specifically, this includes:

[0071] S11, AIS dynamic and static information merging:

[0072] After decoding, AIS information yields dynamic and static information data tables. Because AIS data may contain missing static information and to facilitate subsequent trajectory reconstruction, the dynamic and static information data need to be merged. The specific process is as follows: The static data table is denoted as A, and the dynamic data table as B; using the MMSI attribute column as an index, the attribute columns in table A are matched to those in table B, merging to form a new trajectory data table C; Attribute columns related to ship motion are retained, specifically including MMSI, recording time, longitude, latitude, heading, speed, and ship length information.

[0073] S12. Trajectory data sorting:

[0074] Since the original AIS dynamic data is sorted according to the data reception time, the AIS information is disorganized. Therefore, it is necessary to sort the trajectory data table C according to the MMSI and recording time attribute columns to obtain data reflecting the continuous spatiotemporal movement of the ship.

[0075] At this point, the data structure of the ship trajectory set is defined as: T = {MMSI, Time, Lon, Lat, COG, SOG, Length}. Where MMSI represents the ship's maritime movement business identity, Time represents the recording time, Lon represents the longitude, Lat represents the latitude, COG represents the heading, SOG represents the speed, and Length represents the ship's length. An example table of the ship trajectory set data is shown below:

[0076]

[0077] S2. Based on the ship's speed value, delete ship stop points from the trajectory data table. Using a sliding window algorithm, divide the massive number of trajectory points into different voyage trajectory segments based on the time interval between adjacent points. The specific method is as follows:

[0078] A ship needs to complete a voyage from port A to port B to complete its transportation mission. During adjacent voyages, the ship needs to remain stationary at its berth for a period of time to load and unload cargo, during which time its speed is 0. Therefore, the massive trajectories can be divided into different voyage trajectory segments based on the time interval between adjacent trajectory points. First, delete the data rows in the ship trajectory dataset where the speed is equal to 0; then, extract the trajectory of the same ship based on the MMSI (Modular Missile Indicator), calculate the time difference between adjacent trajectory points using the sliding window algorithm, and determine whether the two points belong to the same voyage based on the dwell time threshold Δt; repeat the voyage segmentation operation in the previous step for all MMSIs to achieve the segmentation of all voyages in the massive trajectory dataset.

[0079] At this point, the data structure of the ship voyage trajectory dataset is: T i,j={MMSI,Time,Lon,Lat,COG,SOG,Length,Voyage}. Where i represents the ship's MMSI, j represents the voyage to which this trajectory belongs, and Voyage represents the number of voyages. Assuming the dwell time threshold Δt is 0.5h, the example table of the trajectory ship trajectory set in step S1 can be converted into two ship voyage trajectories, as shown in the example table below:

[0080]

[0081]

[0082] S3. Identify and remove outliers in position, heading, and speed in each voyage trajectory segment;

[0083] Because data transmission is affected by weather, traffic flow, and human factors, there are outliers in the trajectory position, heading, or speed data that significantly deviate from the normal movement trend. Therefore, it is necessary to identify and delete these outliers. The specific process is as follows:

[0084] S31, Location noise identification:

[0085] For each voyage trajectory data T i,j Two adjacent points P n and P n+1 The distance Δd between n,n+1 It is obtained from the following formula:

[0086] Δd n,n+1 =ED(P n ,P n+1 )

[0087] Where ED(·,·) represents the Euclidean distance between two points. The average velocity between two adjacent points... It can be obtained from the following formula:

[0088]

[0089] Where, Δd n,n+1 This represents the time interval between two adjacent points. If the average velocity between two adjacent points... Greater than the maximum speed during this voyage If the value is 1.2 times that of the point, then these two points can be identified as location noise points, and P can be set accordingly. n and P n+1 Two-point deletion. Location anomaly detection, such as... Figure 2 As shown.

[0090] S32, Speed ​​Noise Recognition:

[0091] The International Telecommunication Union (ITU) specifies the range of ship speeds in AIS information as 0-120.3 knots. To eliminate sudden speed changes, SOG at this point is... n Compare the velocities of two points before and after the trajectory point (SOG) n+1 and SOG n-1 If SOG n If the following formula is satisfied, the speed value can be determined to be abnormal, and the speed value can be deleted.

[0092]

[0093] Among them, v max This represents the maximum value described for ship speed in the ITU standard. Speed ​​anomaly identification is as follows: Figure 3 As shown.

[0094] S33, Heading noise identification:

[0095] COG represents the ship's heading relative to the ground. According to ship design specifications and maneuvering characteristics, the maximum yaw diameter d during a turn is directly proportional to the ship's length L, typically taken as 2-4 times the length, i.e., d = k × L, k ∈ [2, 4]. When a ship performs a turn maneuver with speed v and diameter d, the ship's turning angular rate is... The ship starts from track point P i To trajectory point P i+1 Maximum steering angle θ max It can be obtained by integrating the rate over time, as shown in the following formula:

[0096]

[0097] If the angular velocity of the trajectory point relative to the previous point is greater than θ max If the following formula is satisfied, then the trajectory point is an abnormal turning point and the speed value is deleted.

[0098]

[0099] Note that when |SOG i+1 -SOG i If |SOG| is significantly greater than 180, it indicates that the heading has exceeded 360°. i+1 -SOG i |Should be replaced with 360-|SOG i+1 -SOG i |. Course anomaly identification, such as Figure 4 As shown.

[0100] S4. Based on the sliding window algorithm, the trajectory points in the voyage trajectory segment are divided into turning motion state and straight motion state. Adjacent trajectory points in the same motion state are merged into a straight motion segment or a turning motion segment. The specific process is as follows:

[0101] Based on whether a ship is in a steering maneuver, its motion can be divided into straight-line motion and turning motion. Straight-line motion indicates that the ship is moving steadily along a certain heading with uniform acceleration; its trajectory can be approximated as a straight line. Turning motion indicates that the ship is continuously changing direction by altering the rudder angle; its trajectory can be considered as an arc. Therefore, a sliding window is used to calculate the rate of change of heading between two adjacent points to determine whether the ship is currently in a turning or straight-line motion. Then, adjacent trajectory points in the same motion state are merged into a straight-line motion segment or a turning motion segment. Specific operation method:

[0102] For each voyage trajectory data T i,j Two adjacent points P n and P n+1 The heading difference ΔCOG between n,n+1 It can be obtained from the following formula:

[0103] ΔCOG n,n+1 =|COG n+1 -COG n |

[0104] Among them, COG n+1 COG n Representing P respectively n+1 and P n The heading value of P. n Rate of change of heading at point Δrot n It can be obtained from the following formula:

[0105]

[0106] When Δrot n Greater than the heading change rate threshold θ rot Then, it can be determined that P n The point represents the turning motion state. Otherwise, it is in a straight-line motion state.

[0107] When obtaining the motion state of each point in the voyage trajectory, that is, the motion state of each point in the voyage trajectory Where n = 1, 2, ..., k, indicating that there are k trajectory points in this voyage, and m = {s, t}, where s represents the straight-ahead state and t represents the turning state. Then, when the sliding window algorithm is used to identify adjacent trajectory points that share the same motion state, the adjacent trajectory points are merged into the same motion trajectory segment. At this point, the trajectory of this voyage can be divided into motion segments under different motion states. Where x represents the number of movement segments that the voyage can be divided into, and y represents whether the movement segment is a straight movement segment or a turning movement segment.

[0108] S5. Based on the sliding window algorithm, different motion segments are identified. A specific linear interpolation method or cubic spline interpolation method is adaptively selected based on the motion state characteristics of each motion segment to reconstruct the trajectory segment. Each reconstructed motion segment is then merged into a complete flight trajectory segment. The specific process is as follows:

[0109] Once a voyage trajectory comprising different motion segments is obtained, in order to accurately describe the ship's true motion state, a specific method needs to be selected based on its motion characteristics to interpolate and reconstruct the motion segment trajectory. Before starting the interpolation and reconstruction work, the sliding window algorithm needs to be used to count the number of trajectory points contained in different motion segments within each voyage trajectory. If the number of trajectory points in a motion segment is less than 3, it can be considered as interference motion caused by wind, waves, or current. In this case, the motion segment is merged with the previous motion segment, ultimately obtaining different motion segments with distinct motion state characteristics.

[0110] For flight trajectories containing different motion segments, a sliding window algorithm is used to identify and obtain complete motion segments. An interpolation method is adaptively selected to reconstruct the motion segments based on their motion state characteristics. For straight-line motion segments, a linear interpolation method is used to reconstruct the trajectory, as shown in the following formula:

[0111]

[0112] For the turning motion segment, cubic spline interpolation is used to reconstruct the trajectory, as shown in the formula below. Finally, all reconstructed motion segments within the voyage are merged to form a complete trajectory.

[0113]

[0114] Where t is the reconstruction time. and These are the coefficients of the linear interpolation function. and These are the coefficients of the cubic spline interpolation function. The longitude, latitude, heading, and speed information in the ship trajectory data can all be repaired using the above methods. The reconstructed ship trajectory, speed, and heading results for a certain voyage are shown below. Figure 5 , 6As shown in Figures 7 and 8, the original trajectories and reconstructed results of a large number of ships are respectively as follows: Figure 8 , 9 As shown.

[0115] The results show that the method designed in this invention can process massive amounts of ship trajectories, identify outliers in the trajectories, and reconstruct the trajectories by adaptively selecting interpolation methods based on the ship's motion state, thus providing reliable ship navigation information for maritime traffic knowledge mining.

[0116] The present invention also provides a storage medium comprising a stored program, wherein, when the program is executed, a method for adaptive reconstruction of massive ship trajectories taking into account the ship's motion state is performed.

[0117] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes an adaptive reconstruction method for massive ship trajectories that takes into account the ship's motion state through the computer program.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive reconstruction method for massive ship trajectories considering ship motion states, characterized in that, Includes the following steps: S1. Merge the dynamic and static data in the Automatic Identification System (AIS) data into a trajectory data table, and sort the information in the trajectory data table in ascending order using the MMSI and the record time attribute column. S2. Based on the ship's speed value, delete the ship's stop points in the trajectory data table after ascending sorting. Based on the sliding window algorithm, divide the points in the trajectory data table whose time interval is greater than the time threshold into different voyage trajectory segments. S3. Identify and remove outliers in position, heading, and speed in each voyage trajectory segment; S4. Based on the sliding window algorithm, the trajectory points in the voyage trajectory segment processed by S3 are identified as turning motion state and straight motion state. Adjacent trajectory points with the same motion state are merged into a straight motion segment or a turning motion segment. S5. Based on the sliding window algorithm, identify different motion segments, and adaptively reconstruct the trajectory segment using a specific linear interpolation method or cubic spline interpolation method according to the motion state characteristics of each motion segment. Then, merge each reconstructed motion segment into a complete flight trajectory segment. Specifically, this includes the following steps: The sliding window algorithm is used to count the number of trajectory points contained in different motion segments within each voyage trajectory. If the number of trajectory points in a motion segment is less than 3, the motion segment is merged with the previous motion segment to finally obtain different motion segments with obvious motion state characteristics. For flight trajectories containing different motion segments, specific interpolation methods are selected to reconstruct the motion characteristics. For these trajectories, a sliding window algorithm is used to identify and obtain complete motion segments, and an interpolation method is adaptively selected to reconstruct the segments based on their motion state characteristics. For straight-line motion segments, a linear interpolation method is used to reconstruct the trajectory, as shown in the following formula: For the turning motion segment, the trajectory is reconstructed using cubic spline interpolation, as shown in the following formula; finally, all reconstructed motion segments within this voyage are merged to form a complete trajectory; in, t It is time to rebuild. and These are the coefficients of the linear interpolation function. , , and These are the coefficients of the cubic spline interpolation function.

2. The adaptive reconstruction method for massive ship trajectories considering ship motion states according to claim 1, characterized in that, S1 specifically includes the following steps: S11. Merging dynamic and static data information: After decoding the data from the Automatic Identification System (AIS), dynamic and static data information tables are obtained; the static data table is denoted as A, and the dynamic data table as B; MMSI The attribute columns are used as indexes to match the attribute columns in table A to table B, and then merge them to form a new trajectory data table C; the attribute columns related to ship motion are retained. S12. Trajectory data sorting: According to trajectory data table C MMSI The data is sorted by the time attribute column to obtain data reflecting the continuous spatiotemporal movement of the ship.

3. The adaptive reconstruction method for massive ship trajectories considering ship motion states according to claim 2, characterized in that, In S11, the attributes related to ship motion specifically include: MMSI Record time, longitude, latitude, heading, speed, and ship length information.

4. The adaptive reconstruction method for massive ship trajectories considering ship motion states according to claim 1, characterized in that, S2 specifically includes the following steps: The trajectory is divided into different voyage trajectory segments based on the time interval between adjacent trajectory points. First, data rows with a central speed of 0 are deleted from the ship trajectory dataset; then, based on... MMSI The trajectory of the same ship is extracted, and the time difference between adjacent trajectory points is calculated using a sliding window algorithm, based on a dwell time threshold. Determine whether the two points belong to the same voyage; for all MMSI The previous flight segmentation operation is repeated to segment all flights in the trajectory dataset.

5. The adaptive reconstruction method for massive ship trajectories considering ship motion states according to claim 1, characterized in that, S3 specifically includes the following steps: S31, Location noise identification: For each voyage trajectory data two adjacent points and Distance between It is obtained from the following formula: in, Represents the Euclidean distance between two points and the average velocity between two adjacent points. It is obtained from the following formula: in, This represents the time interval between two adjacent points, if the average velocity between two adjacent points... Much greater than the highest speed during this voyage Then these two points are determined to be location noise points, and... and Two deletions; S32, Speed ​​Noise Recognition: This point Compare the velocities of two points before and after the trajectory point. and ,like If the following formula is satisfied, the speed value can be determined to be abnormal, and the speed value can be deleted. in, This represents the maximum value described in the ITU for ship speed; S33, Heading noise identification: If the turning angular velocity between the trajectory point and the previous point is greater than If the following formula is satisfied, then the trajectory point is an abnormal turning point, and the speed value is deleted; in, For the maximum steering angle, v For ship speed, L For the length of the ship, , Let the initial trajectory point be... This is the final trajectory point.

6. The adaptive reconstruction method for massive ship trajectories considering ship motion states according to claim 1, characterized in that, S4 specifically includes the following steps: The ship's motion state is divided into straight-line motion state and turning motion state. The straight-line motion state means that the ship is stable in a certain heading and undergoes uniform speed change motion. The turning motion state means that the ship achieves continuous change of direction by changing the steering gear angle. The rate of change of heading between two adjacent points is calculated using a sliding window to determine whether the ship is currently in a turning or straight-line state. Then, adjacent trajectory points in the same motion state are merged into a straight-line motion segment or a turning motion segment. For each voyage trajectory data two adjacent points and The difference in heading between It can be obtained from the following formula: in, represent The heading value, represent The heading value, Rate of change of heading at a point It can be obtained from the following formula: when Greater than the heading change rate threshold Then it can be determined The point represents the turning motion state. Otherwise, it is in a straight-line motion state. ; Motion state of each point in the voyage trajectory ,in, This indicates that there are a total of [number] voyages on this voyage. k A trajectory point, , s Indicates direct flight status. t Indicates the turning state; When the sliding window algorithm is used to identify adjacent trajectory points as belonging to the same motion state, the adjacent trajectory points are merged into the same motion trajectory segment. This allows the trajectory of a flight to be divided into motion segments under different motion states. ,in, x This indicates the number of movement segments that can be divided into for this voyage. y This indicates whether the motion segment is a straight-line motion segment or a turning motion segment.

7. A storage medium, characterized in that, The storage medium includes a stored program, wherein when the program is executed, it performs the adaptive reconstruction method for massive ship trajectories that takes into account the ship's motion state as described in any one of claims 1 to 6.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the adaptive reconstruction method for massive ship trajectories that takes into account the ship's motion state, as described in any one of claims 1 to 6, through the computer program.

Images