Dynamic anchor position detection method based on density clustering algorithm optimization

Abstract

The invention provides a dynamic anchor position detection method based on density clustering algorithm optimization. The method comprises the following steps: data cleaning: decoding, abnormal value filtering and standardization processing are carried out on ship AIS data; dBSCAN track point clustering: identifying a ship anchoring and gathering area based on a density clustering algorithm; post-processing a clustering result: filtering noise points, and selecting a cluster containing the most data points as an anchor area; and anchor location point positioning: calculating a coordinate mean value of all points in the anchor location area, and outputting a real anchor location point through anti-standardization. Intelligent transition from a fuzzy'anchoring area 'to an accurate'physical anchoring position point' is realized, the anchoring position can be positioned more accurately, environmental change can be adapted more dynamically, data noise can be resisted effectively, and a real and reliable technical basis is provided for intelligent anchoring monitoring, collision risk early warning and fine management of an anchoring ground.

Description

Technical Field

[0001] This invention relates to the field of ship anchoring operation technology, specifically to a dynamic anchor position detection method based on density clustering algorithm optimization. Background Technology

[0002] In recent years, with the development of intelligent shipping technology, anchorage selection methods have been continuously optimized. For example, existing technologies have disclosed improved anchorage radius models and safety distance models, achieving automated anchorage detection for vessels moored at a single anchor for the first time; other scholars have further improved anchorage detection algorithms by introducing water depth factors and ship type differentiation modeling. Additionally, existing technologies have disclosed that combining Monte Carlo simulation with traversal algorithms significantly improves the accuracy and computational efficiency of anchorage detection.

[0003] A traditional method is typically based on a fixed "anchoring circle" model. The radius of this circle is determined by the actual distance between ships. It is determined that the latter is calculated using the following formula based on parameters such as ship length, water depth, and chain length:

[0004] Among them, point ( , ( ) is the location of an existing vessel or other object obstructing anchoring operations in the Cartesian coordinate system where the anchorage is located; point ( , (This refers to meeting the safe distance for anchored vessels) The anchorage point. Calculated statically. Once this circle is calculated, it's assumed the ship will move within it.

[0005] However, these methods still have obvious limitations: static models are difficult to adapt to dynamic environmental changes, have insufficient conflict detection capabilities when multiple ships are anchored, and have high computational complexity, resulting in poor real-time performance.

[0006] The foregoing background information is intended to help those skilled in the art understand prior art that is similar to the present invention, and to facilitate the understanding of the inventive concept and technical solution of the present invention. It should be clearly stated that, in the absence of clear evidence that the above content was disclosed before the filing date of this patent application, the foregoing background information should not be used to evaluate the novelty of the technical solution of this application. Summary of the Invention

[0007] Technical issues To address the aforementioned problems, the present invention aims to provide a dynamic anchorage detection method optimized using the density clustering algorithm (DBSCAN). This method achieves an intelligent leap from vague "anchorage area" to precise "physical anchorage point," enabling more accurate anchorage location, dynamic adaptation to environmental changes, and effective resistance to data noise. It provides a truly reliable technical foundation for intelligent anchorage monitoring, collision risk warning, and refined anchorage management.

[0008] Technical solution To achieve the above objectives, the inventors of this application conducted in-depth research and discovered that by using density clustering to identify the actual drift aggregation area of ​​a ship in an anchored state, and based on the spatial distribution characteristics of trajectory points within the cluster, including the density of core points and the mean coordinates of points within the cluster, the anchor point can be accurately located. This achieves more accurate anchor point prediction than the traditional static geometric circle center method, which is more consistent with the actual drift behavior of the ship.

[0009] That is, the present invention is: (1) A dynamic anchor detection method based on density clustering algorithm optimization, comprising the following steps: Data cleaning: Decoding, outlier filtering, and standardization of ship AIS data; DBSCAN trajectory point clustering: Identifying ship anchorage clustering areas based on density clustering algorithm; Post-processing of clustering results: Filter out noisy points and select the cluster containing the most data points as the anchor region; Anchor point location: Calculate the average coordinates of all points within the anchor point area, and output the true anchor point after inverse standardization.

[0010] Furthermore, the outlier filtering includes: Speed ​​anomaly filtering: Track points corresponding to anchored vessels with speeds of 0.5-1 knots, track points with sudden speed jumps between 0 and 2 knots or more, and track points with speeds greater than 2 knots for 3 consecutive track points. Location outlier filtering: Delete trajectory points whose adjacent point displacement is greater than 0.1 nautical miles; Time outlier filtering: If the time of the later trajectory point is earlier than the time of the previous trajectory point, it is judged as a time outlier. If the time difference between the two trajectory points is ≤30 minutes, the data is completed by linear interpolation between the two points. Data with a timestamp interruption >30 minutes is marked as lost and no interpolation is performed.

[0011] Furthermore, the speed outlier filtering includes: For vessels at anchor, with a speed of 0.5-1 knots, they drift slightly due to wind and current; this should be removed. If the speed suddenly changes from 0 to 2 or in the opposite direction, it is identified as a fault point and deleted. If the speed of three consecutive trajectory points is greater than 2 knots, the points are identified as false alarms and deleted.

[0012] Furthermore, the standardization process uses the following formula to obtain standardized coordinates. : (1) in These are the original coordinates; The mean; Standard deviation. Original coordinates. Including the original Longitude coordinates in the data and dimensional coordinates .

[0013] Furthermore, the DBSCAN trajectory point clustering specifically includes: For a given point ,That Neighborhood The calculation formula is as follows: (2) Includes all points whose distance to p is less than 0.5, using the following Euclidean distance formula: (3) in, The neighborhood radius; The set of trajectory points to be clustered; for Point and Euclidean distance between points; for The x-coordinate of the point; for The ordinate of the point; for The x-coordinate of the point; for The ordinate of a point. A point is considered a core point when its ε-neighborhood contains at least 10 points (MinPts = 10). Points connected by density, i.e., a series of reachable core points exist between two points, represent the core location of a densely anchored area, exhibiting significant density characteristics. If a point is marked as a core point, it indicates that this area is an effective anchorage cluster.

[0014] Furthermore, in the DBSCAN trajectory point clustering: Neighborhood radius Set to 0.5 nautical miles; The criteria for determining a core point are that its neighborhood contains at least 10 trajectory points.

[0015] Furthermore, the post-processing of the clustering results specifically includes: Filter the discrete points marked as noise; Count the number of data points in each cluster and select the cluster with the most data points as the anchor region.

[0016] Furthermore, this also includes building anchoring areas: Geometric boundaries are extracted from the effective anchoring trajectory clusters in the anchorage area, and a minimum convex polygon containing all extreme trajectory points is constructed. The anchoring domain is then constructed based on the actual drift behavior of the ship.

[0017] Furthermore, the inverse normalization of the anchor point location specifically includes: transforming the mean coordinates through inverse normalization, i.e., formula (1), to map the normalized coordinates back to the true latitude and longitude space to output the true anchor point.

[0018] Furthermore, this also includes anchor point applications: Based on the real-time calculation of the distance to surrounding vessels at the anchor point, if the distance is less than the sum of the safety radii of the two vessels, a collision warning is triggered. Optimize anchorage planning based on the actual drift range to improve the utilization rate of anchorage space.

[0019] According to this invention, the ship's AIS data is first standardized, then clustered using DBSCAN trajectory points to obtain effective anchorage clusters. After filtering out noisy discrete points from the clustering results, the cluster containing the most data points is selected as the main anchorage area. The anchorage center point is determined by calculating the mean coordinates of all points in this cluster. Finally, the standardized coordinates are mapped back to the true latitude and longitude space through inverse standardization transformation to output the true anchorage point. In the post-clustering processing stage, a triple mechanism of "core point identification + maximum cluster selection + density-weighted positioning" is used to achieve an intelligent leap from the fuzzy "anchorage area" to the precise "physical anchorage point". This process is not a simple direct application of clustering results, but rather a deep information mining that integrates the semantics of ship motion state and spatial density distribution characteristics. Compared with traditional methods, this scheme can not only locate the anchorage more accurately, but also dynamically adapt to environmental changes and effectively resist data noise, providing a truly reliable technical foundation for intelligent anchorage monitoring, collision risk warning, and refined anchorage management.

[0020] (2) A dynamic anchor position detection system for implementing the aforementioned method, comprising: The data acquisition module is used to acquire dynamic and static AIS data of the ship; The data processing module is used to perform data cleaning, DBSCAN clustering, anchor point location algorithm, and collision risk assessment. The output module is used to display anchor points and collision risk assessment results in real time.

[0021] Furthermore, the data processing module is also configured as follows: When the vessel drifts beyond a preset threshold range, an anchor drag warning signal is triggered. Aggregating multiple ship anchorages generates an anchorage heat map to assist port scheduling decisions.

[0022] (3) A computer device, the computer device including a memory, a processor, a communication interface and a communication bus; wherein the memory, the processor and the communication interface communicate with each other through the communication bus; the memory is used to store a computer program; the processor is used to execute the computer program stored in the memory, and when the processor executes the computer program, it implements the steps of the aforementioned dynamic anchor detection method based on density clustering algorithm optimization.

[0023] (4) A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the steps of the aforementioned dynamic anchor detection method based on density clustering algorithm optimization.

[0024] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined to obtain specific implementation methods.

[0025] Beneficial effects To address the problem that traditional static models cannot adapt to dynamic environmental changes, this invention achieves adaptive adjustment of anchoring parameters through real-time analysis of ship dynamic data. To address the issue of poor real-time performance due to high computational complexity, this invention significantly improves computational efficiency through optimized algorithm structure. By intelligently analyzing ship dynamic data, this invention effectively solves the shortcomings of traditional methods in terms of environmental adaptability and real-time performance, significantly improving the accuracy and reliability of anchor position detection.

[0026] When a ship drags anchor, meaning its anchor fails to hold onto the seabed, it drifts uncontrollably under the influence of wind and currents. Traditional methods involve crew members manually locating the ship periodically and judging whether its position exceeds the expected drift range, which is inefficient and prone to oversight. In contrast, this invention automates and precisely locates the anchor position. Once the ship knows its exact anchor point, it can accurately calculate the distance to the anchor points of other nearby vessels. By combining the safety radii of the two ships, collision risk can be assessed in real time, triggering an early warning of dragging anchor. If the distance is too close, even before actual displacement occurs, the system can still issue an early warning, indicating the risk.

[0027] Traditional anchorage planning, for safety reasons, allocates a large "anchorage circle" to each ship, resulting in significant space waste. The method of this invention can accurately detect the anchorage position and actual drift range of each ship, allowing port dispatch centers to plan anchorages more scientifically and compactly, enabling more ships to anchor safely within the limited anchorage while ensuring absolute safety. For VTS or port authorities, it provides precise anchorage positions and real-time status of all ships in the entire anchorage, greatly improving management efficiency and emergency response capabilities.

[0028] For intelligent ships, autonomous anchoring and unanchoring are essential functions. This requires the ship to know its anchor and hull position relative to the seabed in real time and with precision, enabling it to make correct decisions. Extensive anchoring data can be used to analyze the holding characteristics of different seabed types and the drift patterns of different ship types under varying wind and wave conditions, thereby optimizing anchoring models to make them more intelligent and safer, providing crucial technical support for intelligent shipping and autonomous driving.

[0029] The present invention adopts the above-mentioned technical solution to achieve the above objectives, which makes up for the shortcomings of the prior art, is reasonably designed, and is easy to operate. Attached Figure Description

[0030] To make the above and / or other objects, features, advantages and examples of the present invention more apparent and understandable, the accompanying drawings used in the specific embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0031] Figure 1 This is a flowchart of the present invention.

[0032] Figure 2 This is the trajectory map of vessel number 1.

[0033] Figure 3 This is the trajectory map of vessel number 2.

[0034] Figure 4 This is the trajectory map of vessel number 3.

[0035] Figure 5 This is the anchorage trajectory of vessel number 4.

[0036] Figure 6 This is the abnormal trajectory point detection map of vessel No. 1.

[0037] Figure 7 This is the abnormal trajectory point detection map of vessel No. 2.

[0038] Figure 8 This is the abnormal trajectory point detection map of vessel No. 3.

[0039] Figure 9 This is the abnormal trajectory point detection map of vessel No. 4.

[0040] Figure 10 This is a predicted map of the ship's trajectory and anchorage point for vessel number 1.

[0041] Figure 11 This is a predicted map of the ship's trajectory and anchorage point for vessel number 2.

[0042] Figure 12 This is a predicted map of the ship's trajectory and anchorage point for vessel number 3.

[0043] Figure 13 This is a predicted map of the ship's trajectory and anchorage point for vessel number 4.

[0044] Figure 14 This is a map showing the anchorage area of ​​vessel number 1.

[0045] Figure 15 This is a map showing the anchorage area of ​​vessel number 2.

[0046] Figure 16 This is a map showing the anchorage area of ​​vessel number 3.

[0047] Figure 17 This is a map showing the anchorage area of ​​vessel number 4.

[0048] Figure 18 This is a comparison diagram of the anchoring trajectory of vessel No. 1 and the anchoring points of the two models.

[0049] Figure 19 This is a comparison diagram of the anchoring trajectory of vessel No. 2 and the anchoring points of the two models.

[0050] Figure 20 This is a comparison diagram of the anchoring trajectory of vessel No. 3 and the anchoring points of the two models.

[0051] Figure 21 This is a comparison diagram of the anchoring trajectory of vessel No. 4 and the anchoring points of the two models. Detailed Implementation

[0052] Those skilled in the art can refer to the content of this document and appropriately replace and / or modify the process parameters to achieve the desired results. However, it should be particularly noted that all similar replacements and / or modifications are obvious to those skilled in the art and are considered to be included in this invention. The products and preparation methods described in this invention have been described through preferred examples, and those skilled in the art can obviously modify or appropriately change and combine the products and preparation methods described herein without departing from the content, spirit, and scope of this invention to realize and apply the technology of this invention.

[0053] Unless otherwise defined, the technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art to which this invention pertains. This invention uses the methods and materials described herein; however, other suitable methods and materials known in the art may also be used. The materials, methods, and examples described herein are illustrative only and are not intended to be limiting. All publications, patent applications, patent cases, provisional applications, database entries, and other references mentioned herein are incorporated herein by reference in their entirety. In case of conflict, the definitions included in this specification shall prevail.

[0054] Unless otherwise specified, the materials, methods, and examples described herein are exemplary and not limiting. While similar or equivalent methods and materials can be used to implement or test the invention, suitable methods and materials are described herein.

[0055] The technical solutions in the embodiments of the present invention will be clearly and completely described below. 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 are within the scope of protection of the present invention. Furthermore, in the absence of conflict, the embodiments and features in the embodiments of this application can be combined with each other.

[0056] To facilitate understanding of the embodiments of the present invention, the abbreviations and key terms that may be involved in the embodiments of the present invention will first be explained or defined. For undefined abbreviations or key terms, they are all conventionally understood by those skilled in the art.

[0057] AIS data: Automatic Identification System data, is real-time dynamic and static information automatically broadcast by ships through AIS equipment. It is used for ship identification, tracking and collision avoidance at sea, and its core function is similar to a ship's "electronic ID card".

[0058] DBSCAN: Density-Based Spatial Clustering of Applications with Noise.

[0059] VTS: Vessel Traffic Services, is a vessel traffic management system established by maritime authorities to monitor, manage, and coordinate vessel navigation in specific waters (such as ports, straits, and canals) to improve navigation safety and efficiency.

[0060] MMSI: Maritime Mobile Service Identity, is a globally unique ship identification code stipulated by the International Maritime Organization (IMO). It consists of 9 digits and is used in maritime communication and monitoring systems such as AIS (Automatic Identification System) and VTS (Vessel Traffic Service).

[0061] No. 1 anchored vessel: 98 meters long and 16 meters wide.

[0062] No. 2 anchored vessel: 53 meters long and 9 meters wide.

[0063] Anchorage No. 3: 97 meters long and 16 meters wide.

[0064] No. 4 anchored vessel: 108 meters long and 16 meters wide.

[0065] Furthermore, unless otherwise specified, the experimental methods used in the embodiments are conventional methods; the materials and reagents used are commercially available unless otherwise specified. Reagents or instruments whose manufacturers are not specified are all commercially available products. All disclosures and other references mentioned herein are incorporated herein by reference in their entirety.

[0066] The present invention is described in detail below. Example 1: like Figure 1 As shown, a dynamic anchor detection method optimized using DBSCAN is provided, including the following steps: Step A: Data Cleaning First, the raw AIS data is cleaned, and then the cleaned data is standardized. The standardization process uses the z-score standardization method to normalize the latitude and longitude coordinates, eliminating dimensional differences and making data at different scales comparable. The standardization formula is as follows: (1) (4) (5) in, Represents the standardized coordinates; The original coordinates include the longitude and latitude coordinates from the original AIS data; The mean; Standard deviation; These are the standardized longitude coordinates; These are the longitude coordinates from the original AIS data; The average longitude; The standard deviation of longitude; These are the standardized latitude coordinates; These are the latitude coordinates from the original AIS data; The latitude mean; This represents the standard deviation of latitude.

[0067] The method for cleaning the raw AIS data described in step A is as follows: Ship trajectory data, i.e., raw AIS data, is obtained through AIS equipment. This ship trajectory data includes dynamic and static data. The dynamic data includes the ship's latitude and longitude coordinates, speed, status, and timestamp. The static data includes the ship identification number (MMSI), ship name, call sign, ship type, and ship dimensions, including length, width, and draft. The method for cleaning the raw AIS data includes the following steps: Step A1, AIS data decoding: AIS data is typically transmitted in the NMEA 0183 protocol format, and the raw message needs to be decoded into structured data. The specific steps are as follows: Step A11, Message Parsing: First, extract the NMEA statement, such as !AIVDM or !AIVDO, with fields separated by commas. Next, decode the 6-bit ASCII encoded AIS payload data, converting the data into a binary stream. Based on the AIS message type, parse the fields according to the protocol: ① Dynamic data: MMSI, latitude and longitude, speed, heading, UTC time, etc. ② Static data: ship name, call sign, ship type, dimensions, etc.

[0068] Step A12, Coordinate Transformation: Convert the raw AIS latitude and longitude to decimal.

[0069] Step A2, Data Preprocessing: After decoding the AIS data, the extracted data is converted into structured data. At the same time, the data is initially cleaned to remove missing values, duplicate data, and obvious errors to ensure data quality.

[0070] Step A3, Bow Conversion The GPS positioning data provided by the Automatic Identification System (AIS) is typically based on the location of GPS devices installed on the vessel. This location often fails to accurately reflect the actual geographical location of critical parts of the vessel. Accurate vessel position information is crucial for detecting the anchorage of moored vessels. Therefore, the parsed data is transformed using the great circle distance formula. Using the latitude and longitude coordinates recorded by the AIS device, the distance from the bow to the center, and the heading, the latitude and longitude of the bow position are calculated using the following formula: (6) Where ∆lat represents the change in coordinates from the GPS antenna position to the bow position in the north-south direction, and ∆lon represents the change in coordinates from the GPS antenna position to the bow position in the east-west direction. This represents the distance from the antenna to the bow of the ship, where R is the Earth's radius. The heading angle is in radians. This represents the latitude value of the GPS antenna. The bow conversion operation is a preparatory step before anchor position detection. For larger vessels, there may be significant discrepancies between the GPS location and the anchor position. To more accurately locate the anchor point, the GPS latitude and longitude data is first converted to the bow data. The longitude and latitude coordinates in the preprocessed AIS data need to be corrected using ∆lat and ∆lon.

[0071] Step A4, Outlier Detection and Handling: Step A41, Speed ​​Anomaly Detection and Handling Speed ​​anomalies are categorized into three types. The first type applies to anchored vessels: speed not exceeding 0.5-1 knots, with slight drift due to wind and current. The second type involves sudden speed jumps; a rapid change from 0 knots to 2 knots or vice versa indicates an AIS equipment malfunction or transmission error. The third type involves multiple consecutive points on the track showing speeds exceeding 2 knots, indicating a false alarm or the vessel not actually anchored. For any of these three types of speed anomalies, the corresponding track point should be deleted.

[0072] Step A42, Location Anomaly Detection and Handling The trajectory points of an anchored vessel typically drift slightly around the anchorage point, with a radius not exceeding 0.1 nautical miles. If the displacement of an adjacent point exceeds 0.1 nautical miles, it is considered a trajectory anomaly, caused by AIS signal drift or error. These trajectory anomalies are then deleted.

[0073] Step A43: Time Anomaly Detection and Handling Anchorage data should be continuous in time, with uniform timestamp intervals, typically 1-10 minutes between AIS reports. If a later trajectory point's time is earlier than a previous trajectory point's time, it is considered an outlier. Furthermore, if a data point is continuously missing for more than half an hour to an hour, it is considered offline or has lost signal. For short-term outliers, data is completed using linear interpolation between preceding and following points. For long-term outliers, data is marked as interrupted and no interpolation is performed.

[0074] Step B: DBSCAN trajectory point clustering The DBSCAN trajectory point clustering uses a density-based spatial clustering algorithm to identify ship clustering areas. For a given point p, its ε-neighborhood... The calculation formula is as follows: (2) Includes all points whose distance to p is less than 0.5, using the following Euclidean distance formula: (3) in, The neighborhood radius; The set of trajectory points to be clustered; for Point and Euclidean distance between points; for The x-coordinate of the point; for The ordinate of the point; for The x-coordinate of the point; for The ordinate of a point. When a point... Neighborhood A point is considered a core point when it contains at least 10 MinPts. Points connected by density, meaning a series of reachable core points exist between two points, represent the core location of a densely anchored area, exhibiting significant density characteristics. Marking such a point as a core point indicates that this area is an effective anchorage cluster.

[0075] Step C: Delineation of Anchorage Area Using the largest valid cluster selected in step B, i.e., all trajectory points within the anchorage cluster, as the input set P, geometric boundary extraction is performed on set P to construct a minimal convex polygon containing all trajectory points. The vertex set of this polygon is... It represents the spatial limit position that a ship can reach while at anchor.

[0076] (7) The maximum radius of the anchorage area. To predict the coordinates of the anchor center point, Let V be the coordinates of the convex hull vertices. Then, iterate through each vertex in the vertex set V. The Euclidean distance formula was used to calculate the distance between the center point of the anchorage and the predicted anchorage. The distance between them. By comparing all distance values, the maximum distance is selected. Finally, using the anchor center point... Centered on the ellipse, with Draw a closed region for the characteristic radius. This region represents the ship's anchorage area and visually reflects the ship's actual maximum drift range under the current environment.

[0077] Step D: Post-processing of clustering results Post-processing analysis of the DBSCAN clustering results was performed. First, discrete points marked as noise were filtered out; these points represent ship trajectory points rather than anchorage points during navigation. Among all valid clusters, the number of data points in each cluster was counted, and the cluster with the most data points was selected as the primary anchorage area. This area represents the region where ships are most concentratedly anchored and has the highest density characteristic.

[0078] Step E: Ship Anchorage Inspection Post-processing was performed on the DBSCAN clustering results to filter out discrete points marked as noise. Among all valid clusters, i.e., the set containing enough trajectory data points that conform to the physical laws of ship anchoring, the largest cluster containing the most data points was selected as the ship anchorage area, which represents the densest anchorage range for ships. The anchorage center point was determined by calculating the mean coordinates of all points in the cluster, and finally, the standardized coordinates were mapped back to the true latitude and longitude space by inverse standardization transformation, i.e., formulas (1)-(3).

[0079] This paper presents a dynamic anchorage detection method that utilizes DBSCAN clustering and the convex hull algorithm for collaborative optimization. Its core lies in breaking the limitations of traditional geometric models: First, DBSCAN density clustering identifies the actual high-density drift clusters of ships in anchored states from AIS data, and the anchorage point is accurately located based on the density and mean coordinates of the core points within the clusters. Then, the convex hull algorithm is introduced to extract the geometric boundaries of the selected effective anchorage trajectory clusters, constructing a minimum convex polygon containing all extreme trajectory points. Through this "clustering localization + convex hull delimitation" approach, this method can construct the anchorage domain based on the actual drift behavior of the ship, thus achieving more accurate anchorage prediction and risk assessment than the traditional static circle center method.

[0080] according to Figure 1 The process shown provides a detailed prediction of the anchorage locations for four anchored vessels, numbered 1, 2, 3, and 4, as detailed below. Figures 2-21 ,in Figures 2-5 The trajectory diagram of the anchored vessel is shown. Figures 6-9 The diagram shows the detection of abnormal trajectory points of anchored vessels. Figures 10-13 A map showing the predicted anchorage points for anchored ships is provided. Figures 14-17 A map showing the anchorage area of ​​the anchored ship is provided. Figures 18-21 The chart shows a comparison of the ship's anchoring trajectory and the anchoring points of the two models. The specific analysis is as follows.

[0081] Figure 2The plot shows the anchoring trajectory of vessel No. 1 after bow-to-bow conversion. The main operations involved preprocessing the trajectory points recorded by AIS data and simultaneously performing bow-to-bow position conversion. By extracting GPS distance to bow position data from the vessel's static information and combining it with the real-time heading, the original latitude and longitude were mapped to the bow position. The plot shows the trajectory points for the last five minutes of the vessel's navigation before anchoring. Light blue dots represent the original trajectory points and the trajectory points from the ship's GPS; dark blue dots represent the trajectory points converted from GPS to the bow position; light pink dots represent navigation points at the GPS location; and red dots represent navigation points at the bow position.

[0082] Figure 3 The plot shows the anchoring trajectory of vessel No. 2 after bow-to-bow conversion. The main operations involved preprocessing the trajectory points recorded by AIS data and simultaneously performing bow-to-bow position conversion. By extracting GPS distance to bow position data from the vessel's static information and combining it with the real-time heading, the original latitude and longitude were mapped to the bow position. The plot shows the trajectory points for the last five minutes of the vessel's navigation before anchoring. Light blue dots represent the original trajectory points and the trajectory points from the ship's GPS; dark blue dots represent the trajectory points converted from GPS to the bow position; light pink dots represent navigation points from the GPS position; and red dots represent navigation points at the bow position.

[0083] Figure 4 The plot shows the anchoring trajectory of vessel No. 3 after bow-to-bow conversion. The main operations involved preprocessing the trajectory points recorded by AIS data and simultaneously performing bow-to-bow position conversion. By extracting GPS distance data from the vessel's static information and combining it with the real-time heading, the original latitude and longitude were mapped to the bow position. The plot shows the trajectory points for the last five minutes of the vessel's navigation before anchoring. Light blue dots represent the original trajectory points and the trajectory points from the ship's GPS; dark blue dots represent the trajectory points converted from GPS to the bow position; light pink dots represent navigation points at the GPS location; and red dots represent navigation points at the bow position.

[0084] Figure 5 The plot shows the anchoring trajectory of vessel No. 4 after bow-to-bow conversion. The main operations involved preprocessing the trajectory points recorded by AIS data and performing bow-to-bow position conversion. By extracting GPS distance to bow position data from the vessel's static information and combining it with the real-time heading, the original latitude and longitude were mapped to the bow position. The plot shows the trajectory points for the last five minutes of the vessel's navigation before anchoring. Light blue dots represent the original trajectory points and the trajectory points from the ship's GPS; dark blue dots represent the trajectory points converted from GPS to the bow position; light pink dots represent navigation points at the GPS location; and red dots represent navigation points at the bow position.

[0085] Figure 6 The graph shows the outlier detection results for the trajectory points of anchored vessel No. 1, primarily using a Z-score-based statistical analysis method. By calculating the Z-score of the velocity of each trajectory point and the distance to adjacent points, points with a velocity exceeding 0.1 knots were identified as velocity anomalies, as indicated by the red crosses in the graph. Points with a distance exceeding 185 meters between adjacent trajectory points were identified as position anomalies, but no position anomalies were observed for this vessel. This operation effectively identifies the vessel's motion state and positional anomalies during anchoring.

[0086] Figure 7 The graph shows the outlier detection results for the trajectory points of anchored vessel No. 2, primarily using a Z-score-based statistical analysis method. By calculating the Z-score of the velocity of each trajectory point and the distance to adjacent points, points with a velocity exceeding 0.1 knots were identified as velocity anomalies, as indicated by the red crosses in the graph. Points with a distance exceeding 185 meters between adjacent trajectory points were identified as positional anomalies, but no positional anomalies were observed for this vessel. This operation effectively identifies the vessel's motion and positional anomalies during anchoring.

[0087] Figure 8 The graph shows the outlier detection results for the trajectory points of anchored vessel No. 3, primarily using a Z-score-based statistical analysis method. By calculating the Z-score of the velocity of each trajectory point and the distance to adjacent points, points with a velocity exceeding 0.1 knots were identified as velocity anomalies, as indicated by the red crosses in the graph. Points with a distance exceeding 185 meters between adjacent trajectory points were identified as position anomalies, but no position anomalies were observed for this vessel. This operation effectively identifies the vessel's motion state and positional anomalies during anchoring.

[0088] Figure 9 The graph shows the outlier detection results for the trajectory points of anchored vessel No. 4, primarily using a Z-score-based statistical analysis method. By calculating the Z-score of the velocity of each trajectory point and the distance to adjacent points, points with a velocity exceeding 0.1 knots were identified as velocity anomalies, as indicated by the red crosses in the graph. Points with a distance exceeding 185 meters between adjacent trajectory points were identified as position anomalies, but no position anomalies were observed for this vessel. This operation effectively identifies the vessel's motion state and positional anomalies during anchoring.

[0089] Figure 10The diagram shows the anchorage location of vessel No. 1. The main operation performed was DBSCAN trajectory point clustering. With a radius of 0.5 nautical miles, a point near at least 10 other vessel trajectory points was marked as a core point, forming a dense cluster of anchorage areas, and further identifying the vessel's anchorage. In the diagram, yellow circles represent the navigation endpoint, and light blue ship icons indicate the bow direction at each trajectory point, visually reflecting the vessel's direction of travel at that point, aiding in determining the vessel's motion status and locating the anchorage. Blue ship icons mark the trajectory start point, purple ship icons represent the endpoint, and yellow pentagrams represent the final anchorage location predicted through clustering and motion analysis.

[0090] Figure 11 The diagram shows the anchorage location of vessel No. 2. The main operation performed was DBSCAN trajectory point clustering. With a radius of 0.5 nautical miles, a point near at least 10 other vessel trajectory points was marked as a core point, forming a dense cluster of anchorage areas, and further identifying the vessel's anchorage. In the diagram, yellow circles represent the navigation endpoint, and light blue ship icons indicate the bow direction at each trajectory point, visually reflecting the vessel's direction of travel at that point, aiding in determining the vessel's motion status and locating the anchorage. Blue ship icons mark the trajectory start point, purple ship icons represent the endpoint, and yellow pentagrams indicate the final anchorage location predicted through clustering and motion analysis.

[0091] Figure 12 The diagram shows the anchorage location of vessel No. 3. The main operation performed was DBSCAN trajectory point clustering. With a radius of 0.5 nautical miles, a point near at least 10 other vessel trajectory points was marked as a core point, forming a dense cluster of anchorage areas, and further identifying the vessel's anchorage. In the diagram, yellow circles represent the navigation endpoint, and light blue ship icons indicate the bow direction at each trajectory point, visually reflecting the vessel's direction of travel at that point, aiding in determining the vessel's motion status and locating the anchorage. Blue ship icons mark the trajectory start point, purple ship icons represent the endpoint, and yellow pentagrams represent the final anchorage location predicted through clustering and motion analysis.

[0092] Figure 13 The diagram shows the anchorage location of vessel No. 4. The main operation performed was DBSCAN trajectory point clustering. With a radius of 0.5 nautical miles, a point near at least 10 other vessel trajectory points was marked as a core point, forming a dense cluster of anchorage areas, and further identifying the vessel's anchorage. In the diagram, yellow circles represent the navigation endpoint, and light blue ship icons indicate the bow direction at each trajectory point, visually reflecting the vessel's direction of travel at that point, aiding in determining the vessel's motion status and locating the anchorage. Blue ship icons mark the trajectory start point, purple ship icons represent the endpoint, and yellow pentagrams represent the final anchorage location predicted through clustering and motion analysis.

[0093] Figure 14The diagram shows the results of constructing the anchorage area for vessel No. 1. This process, based on the convex hull algorithm, selects the anchorage trajectory points farthest from the anchorage point and uses these as the basis for drawing an ellipse, thus defining the vessel's anchorage area. The meanings of the elements in the diagram are as follows: dark blue dots represent the actual anchorage trajectory, red dots represent the navigation trajectory; green triangles and yellow pentagrams mark the anchorage starting point and predicted anchorage point, respectively; the red elliptical area represents the final determined anchorage area.

[0094] Figure 15 The results of constructing the anchorage area for vessel No. 2 are shown. This process, based on the convex hull algorithm, selects the anchorage trajectory points farthest from the anchorage point and uses these as the basis for drawing an ellipse, thus defining the vessel's anchorage area. The meanings of the elements in the figure are as follows: dark blue dots represent the actual anchorage trajectory, red dots represent the navigation trajectory; green triangles and yellow pentagrams mark the anchorage starting point and predicted anchorage point, respectively; the red elliptical area is the final determined anchorage area.

[0095] Figure 16 The results of constructing the anchorage area for vessel No. 3 are shown. This process, based on the convex hull algorithm, selects the anchorage trajectory points farthest from the anchorage point and uses these as the basis for drawing an ellipse, thus defining the vessel's anchorage area. The meanings of the elements in the figure are as follows: dark blue scatter dots represent the actual anchorage trajectory, red scatter dots represent the navigation trajectory; green triangles and yellow pentagrams mark the anchorage starting point and predicted anchorage point, respectively; the red elliptical area is the final determined anchorage area.

[0096] Figure 17 The results of constructing the anchorage area for vessel No. 4 are shown. This process, based on the convex hull algorithm, selects the anchorage trajectory points farthest from the anchorage point and uses these as the basis for drawing an ellipse, thus defining the vessel's anchorage area. The meanings of the elements in the figure are as follows: dark blue dots represent the actual anchorage trajectory, red dots represent the navigation trajectory; green triangles and yellow pentagrams mark the anchorage starting point and predicted anchorage point, respectively; the red elliptical area is the final determined anchorage area.

[0097] Figure 18 The diagram shows a comparison between the anchoring trajectory of vessel No. 1 and the anchoring points of two different models. In the figure, red dots represent the vessel in a sailing state (Status=0), blue dots represent the vessel in an anchored state (Status=1), green pentagons represent the anchoring points predicted by the DBSCAN model, and yellow dots represent the anchoring points predicted by the static model. The anchoring point coordinates predicted by the traditional static model are (110.497383, 21.088562), while the anchoring point coordinates simulated using the DBSCAN clustering model are (110.497831, 21.088720).

[0098] Figure 19The diagram shows a comparison between the anchoring trajectory of vessel No. 2 and the anchoring points of two different models. In the figure, red dots represent the vessel in a sailing state (Status=0), blue dots represent the vessel in an anchored state (Status=1), green pentagons represent the anchoring points predicted by the DBSCAN model, and yellow dots represent the anchoring points predicted by the static model. The anchoring point coordinates predicted by the traditional static model are (110.416221, 21.184134), while the anchoring point coordinates simulated by the DBSCAN clustering model are (110.416282, 21.184121).

[0099] Figure 20 The figure shows a comparison between the anchoring trajectory of vessel No. 3 and the anchoring points of two different models. In the figure, red dots represent the vessel in a sailing state (Status=0), blue dots represent the vessel in an anchored state (Status=1), green pentagons represent the anchoring points predicted by the DBSCAN model, and yellow dots represent the anchoring points predicted by the static model. The anchoring point coordinates predicted by the traditional static model are (110.515800, 21.093368), while the anchoring point coordinates simulated by the DBSCAN clustering model are (110.515647, 21.093342).

[0100] Figure 21 The diagram shows a comparison between the anchoring trajectory of vessel No. 4 and the anchoring points of two different models. In the figure, red dots represent the vessel in a sailing state (Status=0), blue dots represent the vessel in an anchored state (Status=1), green pentagons represent the anchoring points predicted by the DBSCAN model, and yellow dots represent the anchoring points predicted by the static model. The anchoring point coordinates predicted by the traditional static model are (110.679549, 21.064915), while the anchoring point coordinates simulated by the DBSCAN clustering model are (110.679562, 21.064938).

[0101] Figures 2 to 21 The implementation process of this invention on three ships is fully presented: from Z-score-based anomaly trajectory point cleaning, to DBSCAN density clustering identification of the anchorage core area, and finally, accurate location of the anchor point by the mean of the maximum cluster coordinates, which is then compared and verified with a traditional static model. Calculations show that the average distance between the predicted points of the DBSCAN model and the ship's trajectory points after anchoring is smaller, indicating higher prediction accuracy. The results demonstrate that this method can effectively eliminate noise, adapt to the actual drift behavior of the ship, and predict anchor points closer to the actual cluster center, significantly outperforming traditional methods that rely on fixed formulas or geometric averages. This fully verifies the comprehensive advantages of this dynamic detection method in terms of accuracy, robustness, and physical rationality.

[0102] Example 2: A dynamic anchor position detection system for implementing the method described in the foregoing embodiments is provided, comprising: The data acquisition module is used to acquire dynamic and static AIS data of the ship; The data processing module is used to perform data cleaning, DBSCAN clustering, anchor point location algorithm and collision risk assessment; trigger anchor dragging warning signal when the ship drifts beyond the preset threshold range; and aggregate multiple ship anchor points to generate anchorage heat map to assist port scheduling decisions. The output module is used to display anchor points and collision risk assessment results in real time.

[0103] Example 3: A computer device is also provided, comprising a memory, a processor, a communication interface, and a communication bus; wherein the memory, processor, and communication interface communicate with each other through the communication bus; the memory is used to store computer programs; the processor is used to execute the computer programs stored in the memory, and when the processor executes the computer programs, it implements the steps of the dynamic anchor detection method based on density clustering algorithm optimization described in the foregoing embodiments, and can achieve the same technical effect. To avoid repetition, this embodiment will not repeat the steps.

[0104] Example 4: A computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed, it implements the steps of the dynamic anchor detection method based on density clustering algorithm optimization described in the foregoing embodiments and can achieve the same technical effect. To avoid repetition, this embodiment will not repeat the steps.

[0105] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0106] The conventional techniques described in the above embodiments are existing technologies known to those skilled in the art, and therefore will not be described in detail here.

[0107] The specific embodiments described herein are merely illustrative examples illustrating the spirit of the invention. Those skilled in the art to which this invention pertains may make various modifications or additions to the described specific embodiments or use similar methods to substitute them, without departing from the spirit of the invention or exceeding the scope defined by the appended claims. The different technical features recorded in this invention can be combined partially or entirely to obtain technical solutions that conform to the laws of nature; such technical solutions are essentially already described in this invention.

[0108] Although the present invention has been described in detail and specific embodiments have been cited, it will be apparent to those skilled in the art that various changes or modifications can be made without departing from the spirit and scope of the invention.

[0109] While the foregoing detailed descriptions have shown, described, and pointed out novel features applicable to various embodiments, it should be understood that various omissions, substitutions, and changes may be made to the form and details of the described apparatus or methods without departing from the spirit of this disclosure. Furthermore, the various features and methods described above may be used independently of each other or may be combined in various ways. All possible combinations and sub-combinations are intended to fall within the scope of this disclosure. Many of the foregoing embodiments include similar components, and therefore, these similar components are interchangeable in different embodiments. Although the invention has been disclosed in the context of certain embodiments and examples, those skilled in the art will understand that the invention extends beyond the specifically disclosed embodiments to other alternative embodiments and / or applications, as well as their obvious modifications and equivalents. Therefore, the invention is not intended to be limited to the specific disclosure of the preferred embodiments herein.

[0110] All matters not covered in this invention are common knowledge.

Claims

1. A dynamic anchor point detection method based on density clustering algorithm optimization, characterized in that, Comprising the following steps: Data cleaning: decoding, outlier filtering and standardization of ship AIS data; DBSCAN trajectory point clustering: identifying ship anchoring aggregation area based on density clustering algorithm; Post-processing of clustering results: filtering noise points and selecting the cluster containing the most data points as the anchoring area; Anchoring point positioning: calculating the coordinate mean of all points in the anchoring area and outputting the real anchoring point after inverse standardization.

2. The method of claim 1, wherein: The outlier filtering includes: Speed outlier filtering: for ships in anchoring state with speed of 0.5-1 knots, trajectory points with sudden jump in speed of 0-2 knots or more, and trajectory points with continuous 3 trajectory points speed>2 knots; and / or Position outlier filtering: deleting trajectory points with adjacent point displacement>0.1 nautical miles; and / or Time outlier filtering: if the time of the next trajectory point is earlier than that of the previous point, it is determined as a time outlier point, and if the time of the previous and next points is≤30 minutes, the data is completed based on linear interpolation between the two points; data with timestamp interruption>30 minutes is marked as missing and no difference is made.

3. The method of claim 1, wherein: The speed outlier filtering includes: The standardization process obtains standardized coordinates using the following equation : (1) wherein is the original coordinate; is the mean; is the standard deviation.

4. The method of claim 1, wherein: The DBSCAN trajectory point clustering specifically includes: For a given point which neighborhood The formula for the calculation is as follows: (2) comprising all of Points with a distance less than 0.5, distance metric using the following Euclidean distance formula: (3) wherein, is a neighborhood radius; is a set of trajectory points to be clustered; is the Euclidean distance of a point from a point; is the x-coordinate of a point; is the y-coordinate of a point; is the x-coordinate of a point; is the y-coordinate of a point.

5. The method of claim 4, wherein: The post-processing of clustering specifically includes: Filtering discrete points marked as noise; Counting the number of data points in each cluster and selecting the cluster with the most data points as the anchoring area.

6. The method of claim 1, wherein: The real anchoring point outputted after inverse standardization in the anchoring point positioning specifically includes: transforming the coordinate mean by inverse standardization, i.e., formula (1), to map the standardized coordinates back to the real latitude and longitude space to output the real anchoring point.

7. The method of claim 1, wherein: It also includes anchoring point application: Real-time calculation of distance from the anchoring point to surrounding ships, triggering collision warning if the distance is less than the sum of the safety radii of the two ships; Optimizing anchorage planning according to actual drift range to improve anchoring space utilization.

8. The method of claim 1, wherein: It also includes constructing anchoring field: Extracting the geometric boundary of the effective anchoring trajectory cluster of the anchoring area, constructing the smallest convex polygon containing all extreme trajectory points, and constructing the anchoring field based on the actual drift behavior of the ship.

9. A dynamic anchoring detection system implementing the method of any one of claims 1 to 8, characterized in that, It includes: Data acquisition module for acquiring ship AIS dynamic data and static data; Data processing module for executing data cleaning, DBSCAN clustering and anchoring point positioning algorithm; Output module for real-time display of anchoring point and collision risk assessment results.

10. A storage medium having stored thereon a computer program, characterized in that The program is executed to realize the steps of the method of any one of claims 1-8.

Images