A data analysis method, device and equipment for offshore abutting behavior and a medium

Abstract

The application relates to the technical field of data analysis, in particular to a data analysis method, device and equipment for offshore docking behavior and a medium. The application first acquires ship data and calculates a distance value set from a reference point, then performs spatial clustering based on the distance values to filter out potential docking targets, performs trajectory analysis on the targets, and finally determines the docking behavior. This method introduces a preset reference point as a unified reference for distance calculation, converts the traditional two-by-two pairing distance calculation of ships into single-point distance calculation from the reference point, reduces the calculation complexity, and provides a more accurate data basis through the preprocessing link of spatial clustering, so that the system can realize efficient docking behavior recognition and early warning with low power consumption.

Description

Technical Field

[0001] This application relates to the technical field of data analysis, and in particular to a data analysis method, apparatus, equipment and medium for maritime docking behavior. Background Technology

[0002] Maritime illegal activities (such as human smuggling, illegal mining, etc.) are often accompanied by multiple vessels berthing at high speeds. How to efficiently monitor and continuously track maritime targets with such behavioral patterns, and how to provide early warnings for suspicious targets, has become an important issue in current maritime inspection work.

[0003] Existing ship berthing detection systems employ a distance-based calculation method, calculating the distances between all ship data points within a monitored area in real time and combining this with preset thresholds for judgment. This system identifies berthing behavior by continuously comparing changes in distances between ships.

[0004] However, this full-scale calculation method results in a geometric increase in the amount of ship data within the monitored area, leading to an excessive computational load on the system. This not only requires a large investment of computing resources but also affects the ability to predict and continuously track suspicious docking behavior. This situation needs further improvement. Summary of the Invention

[0005] To address the problem that existing analysis methods result in excessive computational loads, requiring significant computing resources and hindering the ability to predict and continuously track suspicious docking activities, this application provides a data analysis method, apparatus, equipment, and medium for maritime docking activities, employing the following technical solution:

[0006] Firstly, this application provides a data analysis method for maritime berthing behavior, comprising the following steps:

[0007] Obtain vessel data within the regulated area;

[0008] Calculate the set of distance values ​​between each data location and a preset reference point based on the ship data;

[0009] Spatial clustering is performed on the data based on the set of distance values ​​to obtain the data clustering results;

[0010] Trajectory analysis is performed based on the data clustering results to obtain trajectory analysis results;

[0011] The existence of a "hitchhiking" behavior is determined based on the trajectory analysis results.

[0012] By adopting the above technical solution, this application provides a new data analysis method to address the computational resource consumption problem in the detection of berthing behavior at sea and to improve the early warning capability for suspicious targets. In traditional monitoring systems, if there are 100 ships in the monitored area, it is necessary to calculate the distance between 4950 pairs of ships. When the number of ships increases to 1000, the computational workload will increase dramatically to 499,500 pairs. This geometric increase severely restricts the system performance. This application first acquires ship data and calculates the set of distance values ​​with reference points. Then, it performs spatial clustering based on these distance values ​​to filter out potential berthing targets. Finally, it performs trajectory analysis on these targets to determine the berthing behavior. This method introduces a preset reference point as a unified benchmark for distance calculation, transforming the traditional pairwise distance calculation of ships into single-point distance calculation with reference points. This not only reduces the computational complexity from O(n²) to O(n), but also provides a more accurate data foundation through the preprocessing stage of spatial clustering, enabling the system to achieve efficient berthing behavior identification and early warning with lower computational power consumption.

[0013] Optionally, after acquiring vessel data within the regulated area, the method further includes the following steps:

[0014] Preliminary location information for each data point is obtained based on the ship data.

[0015] The preliminary position information is subjected to coordinate system transformation, and the transformed position information is corrected for errors according to a preset coordinate reference.

[0016] The corrected location information is then smoothed, and the smoothed location information is used as the final location information.

[0017] By adopting the above technical solution, in actual monitoring, ship position data is often affected by various factors, resulting in deviations. For example, continuous positioning data of a ship within 10 minutes may show irregular drifting within a 500-meter range, while in reality, the ship is in a relatively fixed position. This data noise can seriously affect subsequent berthing behavior analysis. This application first obtains preliminary ship position information, then performs coordinate system normalization processing and combines it with a preset benchmark for error correction, and finally filters out noise caused by short-term fluctuations through a data smoothing algorithm, thereby obtaining high-quality position information. This not only systematically solves the problem of data accuracy but also provides a more reliable data foundation for subsequent berthing behavior analysis, significantly improving the reliability and accuracy of the entire monitoring system.

[0018] Optionally, the ship data includes:

[0019] Automatic Identification System (AIS) data includes the ship's dynamic information, static information, and navigation information;

[0020] Radar detection data, including the ship's real-time position, speed, and heading information;

[0021] Mobile terminal data, including location information and navigation status information collected by shipborne terminals;

[0022] Obtaining vessel data within the regulated area includes the following steps:

[0023] Acquire data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data of each vessel within the regulated area;

[0024] The data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data are fused according to a preset data credibility weight to obtain fused ship data.

[0025] By adopting the above technical solution, traditional monitoring systems that rely solely on Automatic Identification System (AIS) data may experience data loss due to equipment malfunctions or human intervention. This application first simultaneously acquires the ship's AIS data, radar detection data, and mobile terminal data. Then, it fuses these three types of data according to preset data credibility weights to obtain more complete and reliable ship data. By employing multi-source data fusion technology and setting credibility weights for different data sources to optimize the fusion effect, it not only solves the inherent defects of a single data source but also improves the completeness and accuracy of ship information through data complementarity, laying a solid data foundation for subsequent berthing behavior analysis.

[0026] Optionally, the set of distance values ​​between each data position and a preset reference point is calculated based on the ship data, specifically including the following steps:

[0027] Extract ship data for all targets with different identifiers that are online before the current analysis time;

[0028] Status determination is performed on vessel data appearing within the monitored area at the current analysis time:

[0029] Ship data that did not appear or had disappeared in the previous analysis time but appeared in the regulatory area in the current analysis time will be recorded as the first appearance data;

[0030] Ship data that appeared in the previous analysis time and still appears in the regulatory area in the current analysis time will be recorded as continuously occurring data.

[0031] Ship data that appeared in the previous analysis time but did not appear in the online data of the corresponding target identifier in the current analysis time are recorded as missing data;

[0032] Calculate the initial distance between the first occurrence of the data at the current analysis time and the preset reference point;

[0033] Calculate the distance between the continuously occurring data at the current analysis time and the preset reference point;

[0034] The initial distance value and the consecutively occurring distance values ​​are combined to form a distance value set.

[0035] To address the complexity of dynamic target status determination and distance calculation in maritime vessel monitoring, this application provides a target status determination scheme based on time-series analysis. In traditional monitoring systems, since vessel targets may enter or leave the monitored area at any time, failure to accurately classify and track target status can easily lead to data chaos and wasted computational resources. For example, if a vessel briefly leaves the monitored area and then re-enters, it may be misjudged as a new target and subject to repeated calculations. This application first extracts vessel data for all online targets before the current analysis time. Then, by comparing the data with the data from the previous analysis time, it accurately classifies the target status into three types: initial appearance, continuous appearance, and disappearance. Finally, it calculates the distance values ​​between the initial appearance and continuous appearance data and preset reference points, forming a complete set of distance values. This method introduces a time-series comparison mechanism. Through dynamic tracking and accurate classification of vessel target status, it not only avoids repeated calculations and data chaos but also improves the relevance and accuracy of distance calculation through reasonable state transitions, providing more effective data support for subsequent berthing behavior analysis.

[0036] Optionally, spatial clustering is performed on the data based on the set of distance values ​​to obtain the data clustering results, specifically including the following steps:

[0037] Based on the set of distance values ​​and a preset first distance threshold, the data is divided into a near-distance dataset and a far-distance dataset to obtain the data classification result;

[0038] Based on the data classification results, the annulus is divided according to equal radii to obtain the annulus partitioning results;

[0039] Based on the circular partitioning results, the data is grouped into quadrants to determine the target dataset, which is then identified as the data clustering result.

[0040] By adopting the above technical solution, in traditional monitoring systems, directly performing cluster analysis on all ships is not only computationally intensive but also prone to misclassifying ships that are far apart into the same category. For example, in a busy waterway, two ships 5 meters apart may be misjudged as potential docking targets. This application first divides the data into two datasets, near distance and far distance, based on a preset first distance threshold. Then, it divides the data into annular regions based on the principle of equal radius. Finally, it performs quadrant grouping on the annular basis to determine the final target dataset. Through layer-by-layer screening and refined partitioning, this application not only significantly reduces computational complexity but also fully considers the spatial distribution characteristics of ships, effectively avoiding the problem of misjudgment at long distances and providing a reliable data foundation for accurately identifying potential docking behaviors.

[0041] Optionally, trajectory analysis can be performed based on the data clustering results to obtain trajectory analysis results, specifically including the following steps:

[0042] Extract data point pairs from the data clustering results whose distance is less than a preset second distance threshold;

[0043] Analyze the precession trajectory, speed, and heading of the data point pair to determine whether the data point pair is moving in opposite directions;

[0044] Calculate the trend of distance change between the data point pairs;

[0045] The trajectory analysis result is determined based on the judgment result of the opposing motion and the trend of distance change.

[0046] By adopting the above technical solution, in traditional monitoring systems, relying solely on the distance between ships to determine berthing behavior is prone to misjudgment. For example, if two ships are close in distance but have opposite headings, they may be incorrectly identified as potential berthing targets. This application first extracts data point pairs with a distance less than a preset second distance threshold from the clustering results, and then determines whether they are moving towards each other by analyzing their precession trajectory, speed, and heading, and makes a comprehensive judgment in combination with the distance change trend. This method integrates multiple feature dimensions of ship motion into the analysis framework. Through multi-dimensional analysis and comprehensive judgment of motion characteristics, it not only effectively reduces the misjudgment rate, but also more accurately captures the true berthing intention, providing a more accurate early warning basis for maritime supervision and law enforcement.

[0047] Optionally, determining whether there is any collision behavior based on the trajectory analysis results includes the following steps:

[0048] When the distance between the data point pairs is less than a preset third distance threshold, it is determined to be a proximity warning event, and the area, time and number of data points where the proximity warning event occurs are recorded.

[0049] When the distance between the data point pairs is consistently less than the preset third distance threshold and the duration is greater than the preset time threshold, it is determined to be a suspected docking event, and the area, time and number of data points where the suspected docking event occurred are recorded.

[0050] The analysis of vessel data in the suspected docking incidents includes:

[0051] The time of data collection for the first trajectory point of each ship's trajectory is compared, and the ship with the latest data collection time is determined as the last arriving ship;

[0052] Record the acquisition time and location coordinates of the first trajectory point that finally reaches the ship;

[0053] Obtain the coordinates of trajectory points of other vessels near the time of data collection;

[0054] When the distance between the coordinates of the last arriving vessel and the coordinates of any other vessel is greater than a preset fourth distance threshold, it is determined as a successful docking event, and the area, time and number of data points where the successful docking event occurred are recorded.

[0055] By adopting the above technical solution, this application first identifies approach warning events by setting a third distance threshold. When the duration exceeds the preset threshold, it is upgraded to a suspected berthing event. Then, by analyzing the relative positional relationship between the last arriving vessel and other vessels, and combining it with a preset fourth distance threshold, it finally determines whether it is a successful berthing event. Through the refined classification of events and the spatiotemporal correlation analysis of trajectory points, not only is the accuracy of berthing behavior determination improved, but also complete event tracing is achieved, providing detailed data support for subsequent law enforcement and evidence collection, and significantly enhancing the accuracy and reliability of maritime supervision.

[0056] Secondly, this application provides a data analysis device for maritime docking behavior, comprising:

[0057] The vessel data acquisition module is used to acquire vessel data within the regulated area;

[0058] The distance value set calculation module is used to calculate the distance value set between each data position and a preset reference point based on the ship data.

[0059] The data clustering result acquisition module is used to perform spatial clustering on the data based on the distance value set to obtain the data clustering result;

[0060] The trajectory analysis result acquisition module is used to perform trajectory analysis based on the data clustering results and obtain trajectory analysis results.

[0061] The "leaning behavior determination module" is used to determine whether leaning behavior exists based on the trajectory analysis results.

[0062] Thirdly, this application 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 the computer program to implement the steps of the above-mentioned data analysis method for maritime docking behavior.

[0063] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned data analysis method for maritime docking behavior.

[0064] In summary, this application includes at least one of the following beneficial technical effects:

[0065] 1. This application first acquires ship data and calculates the set of distance values ​​with reference points. Then, it performs spatial clustering based on these distance values ​​to filter out potential docking targets. Finally, it performs trajectory analysis on these targets to determine the docking behavior. This method introduces a preset reference point as a unified benchmark for distance calculation, transforming the traditional pairwise distance calculation of ships into single-point distance calculation with reference points. This not only reduces the computational complexity from O(n²) to O(n), but also provides a more accurate data foundation through the preprocessing step of spatial clustering, enabling the system to achieve efficient docking behavior identification and early warning with lower computing power consumption.

[0066] 2. In traditional monitoring systems, data relying solely on Automatic Identification System (AIS) data may be lost due to equipment malfunctions or human intervention. This application first acquires AIS data, radar detection data, and mobile terminal data from the vessel simultaneously. Then, it fuses these three types of data according to preset data credibility weights to obtain more complete and reliable vessel data. By employing multi-source data fusion technology and setting credibility weights for different data sources to optimize the fusion effect, this not only solves the inherent defects of a single data source but also improves the completeness and accuracy of vessel information through data complementarity, laying a solid data foundation for subsequent berthing behavior analysis.

[0067] 3. In traditional monitoring systems, directly clustering all vessels is not only computationally intensive but also prone to misclassifying distant vessels into the same category. For example, in busy waterways, two vessels 5 meters apart might be misjudged as potential docking targets. This application first divides the data into two datasets, near-distance and far-distance, using a preset first distance threshold. Then, it divides the data into annular regions based on the principle of equal radius. Finally, it performs quadrant grouping on the annular region to determine the final target dataset. Through layer-by-layer screening and refined partitioning, this application not only significantly reduces computational complexity but also fully considers the spatial distribution characteristics of vessels, effectively avoiding the problem of misjudgment at long distances and providing a reliable data foundation for accurately identifying potential docking behaviors. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating a data analysis method for maritime berthing behavior according to an embodiment of this application;

[0069] Figure 2 This is a flowchart illustrating the error correction process in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0070] Figure 3 This is a flowchart illustrating step S100 in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0071] Figure 4 This is a flowchart illustrating step S200 in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0072] Figure 5 This is a flowchart illustrating step S300 in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0073] Figure 6 This is a schematic diagram of data clustering in a data analysis method for maritime berthing behavior according to an embodiment of this application;

[0074] Figure 7 This is a flowchart illustrating step S400 in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0075] Figure 8 This is a flowchart illustrating step S500 in a data analysis method for maritime berthing behavior according to an embodiment of this application.

[0076] Figure 9 This is a schematic diagram of the structure of a data analysis device for maritime docking behavior according to an embodiment of this application;

[0077] Figure 10 This is an internal structural diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0078] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0079] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0080] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0081] Firstly, this application provides a data analysis method for maritime berthing behavior, referring to... Figure 1 It includes the following steps:

[0082] S100, Obtain vessel data within the regulated area.

[0083] In this embodiment, the ship data includes information such as the ship's target identifier, position coordinates, speed, heading, and collection time. This data is typically acquired through multi-source data acquisition devices such as Automatic Identification System (AIS), radar, and video surveillance. The regulatory area refers to a pre-defined specific sea area, which can be a key area such as a port, anchorage, or waterway, and is defined by a closed area formed by connecting multiple geographic coordinate points.

[0084] Specifically, the system pre-establishes a regulatory area database, storing the boundary coordinates and attribute information of different types of regulatory areas. During data acquisition, it first receives AIS information transmitted by ships in real time through a data interface, and simultaneously acquires target echo data from radar. After decoding, filtering, and format conversion of these raw data, spatial relationship calculations are performed with the regulatory area database to filter out ship data located within the regulatory area.

[0085] S200. Calculate the set of distance values ​​between each data position and the preset reference point based on the ship data.

[0086] In this embodiment, the preset reference point is a fixed coordinate point predetermined by the system. It can be the geometric center of the monitored area, the location of an important navigational beacon, or other locations with special significance, such as the center point of a high-incidence area determined based on historical data analysis. The distance value calculation adopts a spherical distance calculation model, taking into account the influence of the Earth's curvature to ensure the accuracy of the distance calculation.

[0087] Specifically, the system establishes a reference point database, storing the coordinates and attribute information of reference points corresponding to different regulatory zones. For a vessel within each regulatory zone, the system calculates the spherical distance from the vessel's current position to the corresponding reference point by calling a preset distance calculation function.

[0088] S300. Perform spatial clustering on the data based on the distance value set to obtain the data clustering results.

[0089] In this embodiment, spatial clustering employs a hierarchical clustering method based on distance thresholds. Multiple preset distance thresholds divide the data into different levels, with data within each level exhibiting similar spatial distribution characteristics. The system pre-establishes a clustering parameter configuration library, storing distance thresholds and clustering parameters for different scenarios.

[0090] Specifically, the system first divides the data into a near-distance group and a far-distance group based on a first distance threshold, and then further spatially partitions the data in the near-distance group. During the partitioning process, the system uses a preset clustering algorithm, such as DBSCAN or a modified K-means algorithm, to automatically cluster the data. For example, ships within 100 meters of the reference point are divided into the near-distance group, and those beyond 100 meters are divided into the far-distance group, with more detailed clustering analysis performed within the near-distance group.

[0091] S400. Perform trajectory analysis based on the data clustering results to obtain the trajectory analysis results.

[0092] In this embodiment, a ship motion characteristic model is pre-established, which includes characteristic parameters of typical ship motion patterns, such as changes in speed, course, and turning angle. The system identifies abnormal motion patterns by matching the actual trajectory with the model features.

[0093] Specifically, the system has established a trajectory feature analysis database, storing feature parameters of various types of ships' normal navigation and berthing behaviors. For each clustered data set, the system extracts the ship's historical trajectory points, calculates feature values ​​such as trajectory curvature and rate of change of speed, and compares them with the parameters in the feature database to determine whether there are any abnormal motion patterns. For example, when a ship's speed suddenly decreases and its course changes frequently, the system will mark that trajectory as an object requiring close monitoring.

[0094] S500: Determine whether there is any boarding behavior based on the trajectory analysis results.

[0095] In this embodiment, the determination of berthing behavior is based on a multi-level early warning mechanism. The system has pre-established a berthing behavior feature database, which stores the spatiotemporal feature parameters and determination rules for different types of berthing behavior. These feature parameters include multiple dimensions such as changes in distance between ships, relative motion state, and dwell time.

[0096] Specifically, the system has established a behavior determination rule base, which includes determination conditions and threshold parameters for various docking scenarios. Based on trajectory analysis results, the system evaluates ship behavior level by level according to preset determination rules, and triggers corresponding level of warnings when specific conditions are met. For example, when the distance between two ships is less than a preset threshold and the duration exceeds a set value, the system will determine whether it is docking behavior based on their relative motion state and historical trajectory characteristics.

[0097] In one embodiment, refer to Figure 2 After obtaining vessel data within the regulated area in step S100, the method further includes the following steps:

[0098] S101. Obtain preliminary location information for each data point based on the ship data.

[0099] In this embodiment, the system pre-establishes a ship type mapping table and establishes corresponding data preprocessing rules based on the data characteristics of different types of ships. The mapping table contains key information such as ship type, data acquisition frequency, equipment type, and data format, which is used to quickly identify and preprocess ship location data from different sources;

[0100] Specifically, after receiving vessel data, the system first queries the mapping table based on the vessel's MMSI number to obtain the corresponding preprocessing rules, and then parses the location information field in the data according to the rules. For example, if a fishing vessel sends AIS data every 10 seconds, the system extracts the latitude and longitude information as preliminary location information, and records the corresponding data quality flags for missing data.

[0101] S102. Perform coordinate system transformation on the preliminary position information, and perform error correction on the transformed position information according to the preset coordinate reference.

[0102] In this embodiment, the system pre-establishes a regional coordinate reference library, storing reference point coordinate information and corresponding correction parameters for different regulatory areas. The reference library includes information such as reference point number, standard coordinate value, area coverage, and correction coefficient, used to achieve unified calibration of location data in different areas;

[0103] Specifically, the system first determines the regulatory area where the vessel is located, queries the corresponding benchmark information and correction parameters for that area, and then performs coordinate transformation and error correction on the preliminary position information. For example, if a benchmark in a certain area is a permanent marker with known coordinates, the system calculates a correction coefficient based on the standard coordinate value of that point and applies it to the correction of the position data of all vessels in the area.

[0104] S103. Perform data smoothing processing on the corrected position information, and use the smoothed position information as the final position information.

[0105] In this embodiment, the system pre-sets a data smoothing parameter library and configures corresponding smoothing processing parameters for different scenarios and data characteristics.

[0106] Specifically, the system selects appropriate smoothing parameters based on the ship's historical trajectory characteristics to smooth the corrected position data. For ships at anchor, the system uses a larger time window and a higher smoothing intensity to effectively filter out noise caused by short-term fluctuations and obtain more stable position information.

[0107] In one embodiment, the ship data includes Automatic Identification System (AIS) data, radar detection data, and mobile terminal data; AIS data includes the ship's dynamic information, static information, and navigation information; radar detection data includes the ship's real-time position, speed, and heading information; and mobile terminal data includes position information and navigation status information collected by the shipborne terminal.

[0108] Reference Figure 3 In step S100, obtaining vessel data within the regulated area specifically includes the following steps:

[0109] S110. Obtain data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data of each vessel within the regulated area.

[0110] S120. Based on the preset data credibility weights, the Automatic Identification System (AIS) data, radar detection data, and mobile terminal data are fused to obtain the fused ship data.

[0111] In this embodiment, the system pre-establishes a data fusion rule base to provide a basis for decision-making in data fusion under different scenarios.

[0112] Specifically, the system first calculates the credibility score of each data source at the current moment based on the scoring criteria in the rule base, and determines the fusion weight accordingly. For example, if a ship's AIS signal is stable and matches its historical trajectory, it is assigned a higher weight; conversely, if radar data exhibits localized interference, its weight is reduced accordingly. Based on the weights of each data source, the system performs weighted fusion of key information such as ship position and speed to generate the final ship data. The fused data retains the advantages of each data source while effectively reducing the impact of errors from a single data source.

[0113] In one embodiment, refer to Figure 4 In step S200, the set of distance values ​​between each data position and the preset reference point is calculated based on the ship data, specifically including the following steps:

[0114] S210. Extract ship data for all targets with different target identifiers that are online before the current analysis time.

[0115] In this embodiment, the system pre-establishes a vessel target status database to record and maintain real-time status information of all vessels within the monitored area. This database includes attribute fields such as vessel identifier, status marker, timestamp, location information, and data quality, and is updated periodically to ensure the real-time nature of the status information.

[0116] Specifically, the system queries the status database for all vessel records marked "online" based on a preset analysis time interval (e.g., 5 minutes). For example, during each analysis, the system extracts vessel data with a status of "online" within the most recent time window, including MMSI number, timestamp, latitude and longitude, etc., to prepare for subsequent status determination.

[0117] S220. Determine the status of vessel data appearing within the monitored area at the current analysis time. The determination is as follows:

[0118] Ship data that did not appear or had disappeared in the previous analysis time but appeared in the regulatory area in the current analysis time will be recorded as the first appearance data;

[0119] Ship data that appeared in the previous analysis time and still appears in the regulatory area in the current analysis time will be recorded as continuously occurring data.

[0120] Ship data that appeared in the previous analysis time but did not appear in the online data of the corresponding target identifier in the current analysis time are recorded as disappeared data.

[0121] In this embodiment, the system pre-establishes a target state transition rule base, defining the state determination logic and transition conditions of the ship target at different time points.

[0122] Specifically, the system compares the ship data at the current analysis time with that at the previous analysis time and makes a status determination based on the rule base. For example, if a ship did not appear at the previous time and is detected for the first time at the current time, the system marks it as "first appearance"; if the ship is continuously detected, it is marked as "continuous appearance"; when the ship data disappears at the current time, it is marked as "disappeared".

[0123] S230. Calculate the initial distance between the first occurrence of the data at the current analysis time and the preset reference point.

[0124] S240. Calculate the distance between the continuously occurring data at the current analysis time and the preset reference point.

[0125] In this embodiment, the system pre-establishes a distance calculation parameter library and configures corresponding distance calculation strategies for ships in different states.

[0126] Specifically, the system selects an appropriate calculation strategy based on the ship's status to calculate the distance between the ship and the reference point for both the first appearance and consecutive appearances. For example, for a ship appearing for the first time, a single-point distance calculation is used; for ships appearing consecutively, historical trajectories can be combined for smoothing calculations to improve the stability of the distance values.

[0127] S250, Combine the initial distance value and the consecutively occurring distance values ​​into a distance value set.

[0128] In this embodiment, the system pre-establishes a distance value management library to store and manage different types of distance calculation results.

[0129] Specifically, the system integrates the calculated initial distance value and consecutively occurring distance values ​​into a distance value set according to a predetermined format. The system adds a status flag and timestamp to each distance value and groups and stores them according to ship identifiers, forming a structured distance value dataset.

[0130] In one embodiment, refer to Figure 5 In step S300, spatial clustering is performed on the data based on the distance value set to obtain the data clustering results. This specifically includes the following steps:

[0131] S310. Based on the set of distance values ​​and the preset first distance threshold, the data is divided into a near distance dataset and a far distance dataset to obtain the data classification result.

[0132] In this embodiment, as Figure 6 As shown, the system pre-establishes a distance classification standard library, which contains several key distance thresholds and their corresponding data classification rules based on regulatory experience. In this embodiment, the system sets the first distance threshold to 100 meters and marks it with a solid-lined circular ring in the coordinate system as the boundary between near and far distance datasets.

[0133] Specifically, the system automatically classifies the acquired ship target data according to the rules in the distance classification standard library based on the distance values ​​to the reference point. When the distance value of the ship target data is less than or equal to the first distance threshold, the system marks it as a solid triangle and classifies it into the near-range dataset; when the distance value is greater than the first distance threshold, the system marks it as a hollow circle and classifies it into the far-range dataset.

[0134] S320. Based on the data classification results, divide the rings according to equal radii to obtain the ring partitioning results.

[0135] In this embodiment, as Figure 6As shown, the system pre-constructs a concentric ring hierarchical model, which sets multiple key monitoring distances and their corresponding ring partitioning rules based on historical monitoring data statistics. Within the monitoring area, the system divides the area into inner ring monitoring zones (marked with thin dashed lines) with intervals of 25 meters, and outer ring monitoring zones (marked with thick dashed lines) with radii of 100 meters, 200 meters, 300 meters, and 500 meters.

[0136] Specifically, the system spatially locates the classified ship target data based on a concentric ring hierarchical model. The system associates each ship target with its corresponding ring number and establishes a mapping relationship between ring numbers and ship targets. For example, the system identifies that some ships in the near-field dataset are located within the 25-meter ring, while others are distributed within the 75-100 meter ring area.

[0137] S330. Based on the circular partitioning results, the data is grouped into quadrants to determine the target dataset, and the target dataset is determined as the data clustering result.

[0138] In this embodiment, the system pre-establishes a quadrant spatial distribution feature library, which contains quadrant distribution patterns of various typical ship activities and their corresponding clustering rules. The system divides the monitoring area into four quadrants—northeast, southeast, southwest, and northwest—using two dashed lines for spatial distribution feature analysis.

[0139] Specifically, based on a quadrant spatial distribution feature library, the system determines the quadrant affiliation of ship targets within each ring and analyzes their spatial distribution characteristics. When the system detects multiple ship targets within a specific ring in the same quadrant, it determines whether they conform to a preset clustering pattern in the feature library based on features such as spatial distribution density and relative distance, thereby identifying the target dataset. For example, if the system detects multiple densely distributed ship targets within a 25-meter ring in a certain quadrant, it classifies it as a suspected target dataset.

[0140] In one embodiment, refer to Figure 7 In step S400, trajectory analysis is performed based on the data clustering results to obtain the trajectory analysis results, specifically including the following steps:

[0141] S410. Extract data point pairs from the data clustering results whose distance is less than the preset second distance threshold.

[0142] In this embodiment, the system pre-establishes a database of vessel close-range behavior features. This database summarizes safe distance parameters for different types of vessels under different sea conditions based on historical regulatory data. The system sets a second distance threshold based on the database and uses this threshold as a benchmark for filtering data point pairs. This threshold is then used to perform preliminary filtering of the data clustering results.

[0143] Specifically, the system employs a sliding time window method to retrieve the position data of all ship targets within each time window. When the distance between any two data points is detected to be less than a second distance threshold, the system extracts and saves the complete feature information of this pair of data points (including timestamps, latitude and longitude, ship identification, etc.) to a temporary data pool.

[0144] S420. Analyze the precession trajectory, speed and heading of the data point pair to determine whether the data point pair is moving in opposite directions.

[0145] In this embodiment, the system pre-constructs a ship motion characteristic analysis model, which includes typical characteristic parameters of ship motion and their correlation rules. The model defines the feature vector of the precession trajectory, the threshold range of speed change, and the angle conditions for heading determination, which are used to identify the characteristics of the relative motion between ships.

[0146] Specifically, the system performs feature analysis on data point pairs in the temporary data pool. First, it calculates the motion vector for each pair of data points, including the velocity vector and the direction vector. Then, the system inputs these vectors into the motion feature analysis model and determines whether they meet the conditions for moving towards each other by calculating parameters such as the vector angle and relative velocity. For example, when the heading angle between two ships is within a preset range and the relative velocity is positive, the system marks them as potential moving towards each other targets.

[0147] S430, Calculate the trend of distance changes between data point pairs.

[0148] In this embodiment, the system pre-establishes a distance change pattern library, which contains distance change characteristics of various typical ship interaction behaviors. The library defines multiple distance change curve templates, each corresponding to a specific ship interaction behavior. The system identifies potential abnormal behaviors by matching real-time distance change curves with the templates.

[0149] Specifically, the system employs time-series analysis to calculate the distance difference between adjacent time points in a temporary data pool, generating a distance change sequence. Then, the system uses a moving average algorithm to smooth this sequence, obtaining a distance change trend curve. The system matches this curve with templates in a pattern library to evaluate whether its change characteristics match the expected behavioral patterns.

[0150] S440. Determine the trajectory analysis results based on the judgment results of the opposite motion and the trend of distance change.

[0151] In this embodiment, the system pre-constructs a comprehensive judgment and decision-making model, which integrates judgment rules from two dimensions: opposite motion characteristics and distance change characteristics. The model employs a weighted scoring mechanism, assigning corresponding weights to different characteristic indicators, and assesses the degree of abnormality in ship behavior by calculating a comprehensive score.

[0152] Specifically, the system inputs the results of the relative motion judgment and the distance change trend analysis into the comprehensive judgment and decision-making model. The model first standardizes each feature indicator, and then calculates a comprehensive score based on preset weights. When the score exceeds a preset threshold, the system marks the data point pair as a target requiring special attention and generates a corresponding trajectory analysis report.

[0153] In one embodiment, refer to Figure 8 In step S500, the determination of whether there is any overlap behavior based on the trajectory analysis results includes the following steps:

[0154] S510. When the distance between data point pairs is less than the preset third distance threshold, it is determined to be a proximity warning event, and the area, time and number of data points where the proximity warning event occurs are recorded.

[0155] In this embodiment, the system pre-establishes a database of vessel approach behavior features. Based on historical regulatory data, this database constructs approach distance threshold models for different types of vessels and under different sea conditions. The database includes a mapping relationship between basic parameters such as vessel type, tonnage, and length and safe distances, used to dynamically adjust a preset third distance threshold.

[0156] Specifically, the system employs a real-time monitoring mechanism, matching the type parameters of the currently monitored vessels against a feature library to obtain a corresponding third distance threshold. When the distance between any pair of data points is detected to be less than this threshold, the system automatically creates an early warning event record and stores the relevant information in the event database. This record contains complete data items such as event ID, occurrence time, geographic coordinates, and information about the vessels involved.

[0157] S520. When the distance between data point pairs is continuously less than the preset third distance threshold and the duration is greater than the preset time threshold, it is determined to be a suspected overlap event, and the area, time and number of data points where the suspected overlap event occurred are recorded.

[0158] In this embodiment, the system pre-constructs a time series analysis model, which includes time-based characteristics of ship interaction behavior. By analyzing historical berthing cases, the model summarizes the typical duration characteristics of different types of berthing behaviors and establishes a dynamic adjustment mechanism for time thresholds.

[0159] Specifically, the system continuously tracks early warning events and uses a sliding time window method to calculate the cumulative duration of close proximity between vessels. When the cumulative duration exceeds a preset time threshold, such as 10 minutes, the system escalates the early warning event to a suspected berthing event.

[0160] S530, Analyze vessel data in suspected docking incidents.

[0161] Step S530 includes:

[0162] The time of data collection for the first trajectory point of each ship's trajectory is compared, and the ship with the latest data collection time is determined as the last arriving ship;

[0163] Record the acquisition time and location coordinates of the first trajectory point that finally reaches the ship;

[0164] Obtain the coordinates of trajectory points of other ships near the time of data collection;

[0165] When the distance between the coordinates of the last arriving vessel and the coordinates of any other vessel is greater than a preset fourth distance threshold, it is determined as a successful docking event, and the area, time and number of data points where the successful docking event occurred are recorded.

[0166] Specifically, the system first performs time-series analysis on the vessel trajectory data in suspected docking events, determining the last arriving vessel by comparing the initial data collection time of each vessel. Then, the system extracts the initial position coordinates of that vessel and retrieves position information for similar time points from the trajectory data of other vessels. Finally, the system calculates the spatial distance relationships between these positions; when the distance exceeds a preset fourth distance threshold, the system marks the event as a successful docking event.

[0167] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0168] Secondly, this application provides a data analysis device for maritime docking behavior. The data analysis device for maritime docking behavior of this application will be described below in conjunction with the above-mentioned data analysis method for maritime docking behavior.

[0169] Reference Figure 9 A data analysis device for maritime docking behavior, comprising:

[0170] The vessel data acquisition module is used to acquire vessel data within the regulated area;

[0171] The distance value set calculation module is used to calculate the distance value set between each data position and a preset reference point based on the ship data;

[0172] The data clustering result acquisition module is used to perform spatial clustering on the data based on the distance value set to obtain the data clustering results;

[0173] The trajectory analysis result acquisition module is used to perform trajectory analysis based on data clustering results and obtain trajectory analysis results.

[0174] The "Aggregation Behavior Determination" module is used to determine whether aggregation behavior exists based on trajectory analysis results.

[0175] In one embodiment, the apparatus further includes:

[0176] The location information acquisition module is used to acquire preliminary location information of various data based on ship data.

[0177] The coordinate transformation module is used to transform the initial position information into a coordinate system and to correct the error of the transformed position information according to a preset coordinate reference.

[0178] The data smoothing module is used to smooth the corrected location information and use the smoothed location information as the final location information.

[0179] In one embodiment, the ship data includes:

[0180] Automatic Identification System (AIS) data includes the ship's dynamic information, static information, and navigation information;

[0181] Radar detection data, including the ship's real-time position, speed, and heading information;

[0182] Mobile terminal data, including location information and navigation status information collected by shipborne terminals;

[0183] The ship data acquisition module includes:

[0184] The data acquisition unit is used to acquire data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data of each vessel within the regulatory area.

[0185] The data fusion unit is used to fuse data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data according to preset data confidence weights to obtain fused ship data.

[0186] In one embodiment, the distance value set calculation module includes:

[0187] The ship data extraction unit is used to extract ship data with different target identifiers that are online before the current analysis time;

[0188] The status determination unit is used to determine the status of ship data that appears within the monitored area at the current analysis time.

[0189] Ship data that did not appear or had disappeared in the previous analysis time but appeared in the regulatory area in the current analysis time will be recorded as the first appearance data;

[0190] Ship data that appeared in the previous analysis time and still appears in the regulatory area in the current analysis time will be recorded as continuously occurring data.

[0191] Ship data that appeared in the previous analysis time but did not appear in the online data of the corresponding target identifier in the current analysis time are recorded as missing data;

[0192] The initial distance calculation unit is used to calculate the initial distance value between the first occurrence of data at the current analysis time and the preset reference point.

[0193] The continuous distance calculation unit is used to calculate the continuous distance between the continuously occurring data at the current analysis time and the preset reference point;

[0194] The distance value set generation unit is used to combine the first-occurrence distance value and the consecutively occurring distance values ​​into a distance value set.

[0195] In one embodiment, the data clustering result acquisition module includes:

[0196] The data classification unit is used to divide the data into near-distance datasets and far-distance datasets based on the set of distance values ​​and a preset first distance threshold, thereby obtaining the data classification results;

[0197] The circular partitioning unit is used to divide the data classification results into annular partitions based on equal radii.

[0198] Quadrant grouping unit is used to group data into quadrants based on the circular partitioning results, determine the target dataset, and identify the target dataset as the data clustering result.

[0199] In one embodiment, the trajectory analysis result acquisition module includes:

[0200] The data point pair extraction unit is used to extract data point pairs whose distance is less than a preset second distance threshold from the data clustering results.

[0201] The opposing motion judgment unit is used to analyze the precession trajectory, speed and heading of the data point pair to determine whether the data point pair is moving towards each other;

[0202] The distance change trend calculation unit is used to calculate the distance change trend between data point pairs;

[0203] The trajectory analysis result determination unit is used to determine the trajectory analysis result based on the judgment result of the opposite motion and the distance change trend.

[0204] In one embodiment, the docking behavior determination module includes:

[0205] The early warning event determination unit is used to determine an approaching early warning event when the distance between data point pairs is less than a preset third distance threshold, and to record the area, time and number of data points where the approaching early warning event occurs.

[0206] The suspected overlap event determination unit is used to determine a suspected overlap event when the distance between data point pairs is continuously less than a preset third distance threshold and the duration is greater than a preset time threshold. It records the area, time and number of data points where the suspected overlap event occurs.

[0207] The successful berthing event determination unit is used to analyze vessel data in suspected berthing events, specifically for:

[0208] The time of data collection for the first trajectory point of each ship's trajectory is compared, and the ship with the latest data collection time is determined as the last arriving ship;

[0209] Record the acquisition time and location coordinates of the first trajectory point that finally reaches the ship;

[0210] Obtain the coordinates of trajectory points of other ships near the time of data collection;

[0211] When the distance between the coordinates of the last arriving vessel and the coordinates of any other vessel is greater than a preset fourth distance threshold, it is determined as a successful docking event, and the area, time and number of data points where the successful docking event occurred are recorded.

[0212] In one embodiment, this application provides an electronic device, which may be a server, and its internal structure diagram may be as follows: Figure 10 As shown, the electronic device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a data analysis method for maritime docking operations.

[0213] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0214] In one embodiment, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0215] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0216] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A data analysis method for maritime docking behavior, characterized in that, Includes the following steps: Obtain vessel data within the regulated area; Calculate the set of distance values ​​between each data location and a preset reference point based on the ship data; Spatial clustering is performed on the data based on the set of distance values ​​to obtain the data clustering results; Trajectory analysis is performed based on the data clustering results to obtain trajectory analysis results; The existence of a "hitchhiking" behavior is determined based on the trajectory analysis results.

2. The data analysis method for maritime docking behavior according to claim 1, characterized in that, After acquiring vessel data within the regulated area, the method further includes the following steps: Preliminary location information for each data point is obtained based on the ship data. The preliminary position information is subjected to coordinate system transformation, and the transformed position information is corrected for errors according to a preset coordinate reference. The corrected location information is then smoothed, and the smoothed location information is used as the final location information.

3. The data analysis method for maritime docking behavior according to claim 1, characterized in that, The ship data includes: Automatic Identification System (AIS) data includes the ship's dynamic information, static information, and navigation information; Radar detection data, including the ship's real-time position, speed, and heading information; Mobile terminal data, including location information and navigation status information collected by shipborne terminals; Obtaining vessel data within the regulated area includes the following steps: Acquire data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data of each vessel within the regulated area; The data from the Automatic Identification System (AIS), radar detection data, and mobile terminal data are fused according to a preset data credibility weight to obtain fused ship data.

4. The data analysis method for maritime docking behavior according to claim 1, characterized in that, The calculation of the set of distance values ​​between each data location and a preset reference point based on the ship data includes the following steps: Extract ship data for all targets with different identifiers that are online before the current analysis time; Status determination is performed on vessel data appearing within the monitored area at the current analysis time: Ship data that did not appear or had disappeared in the previous analysis time but appeared in the regulatory area in the current analysis time will be recorded as the first appearance data; Ship data that appeared in the previous analysis time and still appears in the regulatory area in the current analysis time will be recorded as continuously occurring data. Ship data that appeared in the previous analysis time but did not appear in the online data of the corresponding target identifier in the current analysis time are recorded as missing data; Calculate the initial distance between the first occurrence of the data at the current analysis time and the preset reference point; Calculate the distance between the continuously occurring data at the current analysis time and the preset reference point; The initial distance value and the consecutively occurring distance values ​​are combined to form a distance value set.

5. The data analysis method for maritime docking behavior according to claim 1, characterized in that, Spatial clustering of the data based on the distance value set yields the data clustering results, specifically including the following steps: Based on the set of distance values ​​and a preset first distance threshold, the data is divided into a near-distance dataset and a far-distance dataset to obtain the data classification result; Based on the data classification results, the annulus is divided according to equal radii to obtain the annulus partitioning results; Based on the circular partitioning results, the data is grouped into quadrants to determine the target dataset, which is then identified as the data clustering result.

6. The data analysis method for maritime docking behavior according to claim 1, characterized in that, Trajectory analysis is performed based on the data clustering results to obtain trajectory analysis results, specifically including the following steps: Extract data point pairs from the data clustering results whose distance is less than a preset second distance threshold; Analyze the precession trajectory, speed, and heading of the data point pair to determine whether the data point pair is moving in opposite directions; Calculate the trend of distance change between the data point pairs; The trajectory analysis result is determined based on the judgment result of the opposing motion and the trend of distance change.

7. The data analysis method for maritime docking behavior according to claim 6, characterized in that, Determining whether a collision occurred based on the trajectory analysis results includes the following steps: When the distance between the data point pairs is less than a preset third distance threshold, it is determined to be a proximity warning event, and the area, time and number of data points where the proximity warning event occurs are recorded. When the distance between the data point pairs is consistently less than the preset third distance threshold and the duration is greater than the preset time threshold, it is determined to be a suspected docking event, and the area, time and number of data points where the suspected docking event occurred are recorded. The analysis of vessel data in the suspected docking incidents includes: The time of data collection for the first trajectory point of each ship's trajectory is compared, and the ship with the latest data collection time is determined as the last arriving ship; Record the acquisition time and location coordinates of the first trajectory point that finally reaches the ship; Obtain the coordinates of trajectory points of other vessels near the time of data collection; When the distance between the coordinates of the last arriving vessel and the coordinates of any other vessel is greater than a preset fourth distance threshold, it is determined as a successful docking event, and the area, time and number of data points where the successful docking event occurred are recorded.

8. A data analysis device for maritime docking behavior, characterized in that, include: The vessel data acquisition module is used to acquire vessel data within the regulated area; The distance value set calculation module is used to calculate the distance value set between each data position and a preset reference point based on the ship data. The data clustering result acquisition module is used to perform spatial clustering on the data based on the distance value set to obtain the data clustering result; The trajectory analysis result acquisition module is used to perform trajectory analysis based on the data clustering results and obtain trajectory analysis results. The "leaning behavior determination module" is used to determine whether leaning behavior exists based on the trajectory analysis results.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the data analysis method for maritime berthing behavior as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the data analysis method for maritime berthing behavior as described in any one of claims 1-7.

Images