Ship arrival status determination method, device and readable storage medium

Abstract

The application provides a ship port arrival state determination method, device and readable storage medium. The ship port arrival state determination method comprises the following steps: acquiring sailing dynamic data of a target ship, and determining a sailing track terminal position data of the target ship according to the sailing dynamic data; determining a plurality of candidate ports in a preset sailing range according to the sailing track terminal position data, and acquiring port data of each candidate port; determining a dynamic approaching degree of the target ship relative to each candidate port according to the plurality of port data; performing convergence analysis on the plurality of dynamic approaching degrees, and determining that a candidate port meeting a convergence condition in the plurality of port data is a target port for stopping of the target ship; determining a port arrival confidence degree based on a dynamic approaching degree and derived characteristics corresponding to the target port; and determining a port arrival state of the target ship based on the port arrival confidence degree. Through dynamic approaching degree convergence analysis and random forest modeling, the application realizes accurate identification of a ship under an unknown destination port.

Description

Technical Field

[0001] This invention relates to the field of shipping and maritime monitoring technology, and more specifically, to a method, apparatus, and readable storage medium for determining the arrival status of a ship. Background Technology

[0002] In related technologies, existing methods for determining ship arrival rely on ships reporting their destination port or preset port coordinates. When destination port information is missing, incorrect, or the berthing target of ships in densely populated areas of multiple ports is unclear, the approach index cannot be effectively calculated, resulting in a high false alarm rate, poor adaptability, and inability to predict arrival in advance. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] Therefore, the first aspect of the present invention proposes a method for determining the arrival status of a ship.

[0005] A second aspect of the present invention provides a device for determining the arrival status of a ship.

[0006] A third aspect of the present invention provides another device for determining the arrival status of a ship.

[0007] The fourth aspect of this application proposes a readable storage medium.

[0008] In view of this, a first aspect of the present invention provides a method for determining the arrival status of a vessel, comprising: acquiring navigation dynamic data of a target vessel, and determining the final position data of the target vessel's trajectory based on the navigation dynamic data; determining multiple candidate ports within a preset navigation range based on the final position data of the trajectory, and acquiring port data for each candidate port; determining the dynamic proximity of the target vessel relative to each candidate port based on the multiple port data; performing convergence analysis on the multiple dynamic proximity values ​​to determine the candidate ports among the multiple port data that meet the convergence conditions as the target ports for the target vessel to dock; determining the arrival confidence level based on the dynamic proximity value and derived features corresponding to the target port; and determining the arrival status of the target vessel based on the arrival confidence level.

[0009] The navigation dynamic data provided in this application refers to real-time navigation-related data collected by the target vessel through the Automatic Identification System (AIS), including latitude and longitude, speed, heading, timestamp, etc., which is the core data support reflecting the real-time motion status of the vessel. The track end position data refers to the set of key position information of the target vessel in the later stage of navigation, which is extracted from the navigation dynamic data after preprocessing and is used to clarify the search benchmark for potential port candidates.

[0010] The preset navigation range refers to a geographical search area centered on the end position of the vessel's trajectory, within 50 nautical miles, used to filter potential ports of call. Candidate ports refer to all ports within the preset navigation range, selected through geographical matching to identify potential berthing targets. Port data refers to the core attribute information of each candidate port, including coordinates, name, and boundaries, providing basic coordinates and features for dynamic proximity calculations.

[0011] Dynamic approach degree refers to a normalized index that characterizes the degree of approach of a target ship to a candidate port. The value range is [0,1]. It is calculated by integrating the distance change rate, heading deviation and speed change rate. The larger the value, the more obvious the approach trend.

[0012] Convergence analysis refers to the time series analysis of the dynamic convergence sequences corresponding to multiple candidate ports, and the analysis process of judging whether ships are really approaching a certain port by the trend of mean change and volatility.

[0013] The convergence condition refers to the core criteria for determining whether a candidate port is the actual target port, namely, the mean of dynamic convergence continuously increases and the volatility continuously decreases and tends to stabilize.

[0014] The target port refers to the port that the ship actually berths at, selected from multiple candidate ports, meets the convergence criteria, and has the best convergence characteristics.

[0015] Derived features refer to statistical features and inherent ship attributes derived from dynamic approach and navigation dynamic data, including the mean, volatility, and trend coefficient of dynamic approach, which are used to enrich the calculation dimensions of arrival confidence.

[0016] Arrival confidence level refers to the probability value that represents the possibility of a target vessel arriving at the port. The value range is [0,1]. It is calculated by inputting the dynamic approach degree and derived features into the trained model, providing a quantitative basis for determining the arrival status.

[0017] Arrival status refers to the classification of the target vessel's navigation status, including three categories: arrived, expected to arrive, and not yet arrived, which are determined based on arrival confidence and preset thresholds.

[0018] Based on the above data and steps, the solution proposed in this application addresses the technical problems of relying on ships' self-reported destination ports or preset port coordinates for determining ship arrival. When destination port information is missing, incorrect, or the berthing target in a densely populated area with multiple ports is unclear, the solution cannot effectively identify the ship's true destination port and arrival status, leading to high false alarm rates, poor adaptability, and a lack of advance prediction capabilities. The solution achieves automatic inference of the ship's true destination port and accurate, advance determination of its arrival status under unknown destination port conditions, significantly improving port scheduling efficiency and the level of intelligent shipping management. Specifically, by acquiring the target ship's navigation dynamic data and determining the final position data of its trajectory, it eliminates the need to rely on the ship's self-reported destination port information, locking in potential berthing areas solely based on the ship's actual navigation trajectory, thus overcoming the limitation of dependence on declared destination port information. Next, using the final position data of the trajectory as a benchmark, multiple candidate ports are screened within a preset navigation range, and port data is acquired to construct a comprehensive set of potential berthing targets, avoiding omissions or misjudgments caused by judging a single port. Then, a dynamic approach score is calculated for each candidate port, comprehensively quantifying dynamic behaviors such as changes in distance between the ship and the port, course consistency, and speed adjustments. This allows for accurate capture of the trend of ships approaching a specific port. Subsequently, through convergence analysis of multiple dynamic approach scores, candidate ports with continuously increasing mean and stable volatility are selected as the true target ports. This enables autonomous inference of the destination port from a behavioral pattern perspective, effectively distinguishing between transit and arrival behaviors and solving the challenge of destination port identification in complex navigation scenarios. Next, the arrival confidence score is calculated based on the dynamic approach score and derived features corresponding to the target port, integrating dynamic trends and inherent ship attributes to improve the comprehensiveness and accuracy of arrival probability assessment. Finally, the arrival status is determined based on the arrival confidence score and a preset threshold. This not only clearly identifies the arrival and non-arrival states but also predicts the expected arrival status in advance, providing sufficient preparation time for port scheduling, significantly shortening the determination delay, and overcoming the lag in static rule-based determinations. The entire process forms a complete technical chain of trajectory analysis, candidate generation, proximity calculation, destination port inference, confidence assessment, and state determination, realizing the improvement from relying on the input destination to autonomously judging the destination, and significantly improving the coverage and accuracy of ship arrival identification in the scenario of unknown destination port.

[0019] In some technical solutions of this application, acquiring the navigation dynamic data of the target vessel and determining the final position data of the target vessel's track based on the navigation dynamic data includes: acquiring the navigation dynamic data of the target vessel, removing outliers and drift data from the navigation dynamic data, and organizing the data into track data in chronological order. Based on the track data, the final position data of the target vessel's track is determined.

[0020] In this technical solution, outliers refer to extreme data in navigation dynamics that deviate from normal navigation patterns, such as latitude and longitude jumps or abnormal speed spikes caused by equipment failure or signal interference.

[0021] Drift data refers to offset data in navigation dynamics that does not match the actual navigation trajectory of a ship, such as position drift data caused by AIS signal delay.

[0022] Track data refers to a continuous data sequence that accurately reflects a ship's navigation path, after outliers and drift data have been removed and arranged in chronological order.

[0023] This application generates real and continuous track data by removing outliers and drift data from navigation dynamic data and organizing the data in chronological order. This provides a reliable data foundation for the accurate determination of the final position data of the track and avoids the subsequent candidate port selection bias and destination port inference errors caused by noise in the original data.

[0024] Navigational dynamic data is susceptible to factors such as equipment failure, signal interference, and transmission delays during acquisition, resulting in outliers and drift data. Directly using this data for analysis can distort the trajectory, thus affecting the accuracy of the trajectory's final position data. By acquiring the target vessel's navigational dynamic data and first removing outliers and drift data, data noise can be filtered out, restoring the vessel's true navigation trajectory. Then, the remaining valid data is processed chronologically to form continuous trajectory data, ensuring temporal consistency and guaranteeing accurate extraction of the trajectory's final position data. For example, if a vessel experiences a temporary AIS equipment malfunction during navigation, generating anomaly data with latitude and longitude jumps far from the route, this technical solution's processing steps remove this anomaly data. The remaining data, processed chronologically, forms trajectory data that accurately reflects the vessel's navigation path. The trajectory's final position data determined based on this data can accurately pinpoint key areas in the later stages of the vessel's navigation, ensuring the effectiveness of subsequent candidate port selection.

[0025] In some technical solutions of this application, determining the final position data of the target vessel's track based on track data includes: smoothing the speed and heading data in the track data, standardizing the coordinate, time, and distance units of all data in the track data, and generating standardized track data. The final position data of the target vessel's track is then determined based on the standardized track data.

[0026] In this technical solution, smoothing refers to filtering the speed and heading data in the track data to eliminate high-frequency fluctuations, making the data changes smoother and better reflecting the true trend of ship navigation. Coordinate units refer to units of measurement representing latitude and longitude, such as degrees, minutes, and seconds, and must be standardized to a consistent unit format to ensure the accuracy of distance calculations. Time units refer to units of measurement representing time, such as seconds and minutes, and must be standardized to a consistent unit format to ensure the uniformity of change rate calculations. Distance units refer to units of measurement representing distance, such as meters and nautical miles, and must be standardized to a consistent unit format to ensure the consistency of distance-related indicator calculations. Standardized track data refers to track data after smoothing and unit standardization, possessing the characteristics of stable data, uniform format, and reliable accuracy, and is the core basis for accurately determining the final position data of the track segment.

[0027] This application generates standardized track data by smoothing and unifying the units of speed and heading in track data. This further improves the reliability and consistency of track data, ensuring the accuracy of the calculation of the final position data of the track, and laying a high-precision data foundation for subsequent candidate port generation and dynamic approach calculation. Speed ​​and heading data in track data may exhibit high-frequency fluctuations, such as data jitter caused by slight adjustments to the ship's heading or speed. These fluctuations can affect the judgment of the ship's true navigation trend. Simultaneously, the original dynamic navigation data may have inconsistent units for coordinates, time, and distance, directly affecting the accuracy of subsequent calculations. By smoothing the speed and heading data in track data, meaningless high-frequency fluctuations can be filtered out, highlighting the overall trend of the ship's navigation. By unifying the units for coordinates, time, and distance, all data formats are ensured to be consistent, avoiding calculation errors caused by unit differences. For example, a ship's speed data may fluctuate frequently due to wind and waves, and its heading data may experience slight jitter due to manual fine-tuning. After smoothing, the trends in speed and heading data are more stable, accurately reflecting the ship's navigation intentions. Meanwhile, the distance data in the original data that was in meters and distance data that was in nautical miles were unified into nautical miles to ensure the accuracy of subsequent calculations of spherical distance and distance change rate. The accuracy of the final position data of the track determined based on the standardized track data was significantly improved, which provided strong support for the accurate selection of candidate ports.

[0028] In some technical solutions of this application, the dynamic approach degree of the target vessel relative to each candidate port is determined based on multiple port data, including: calculating the spherical distance at consecutive time points using the semi-versus formula based on the vessel's latitude and longitude and the candidate port coordinates; determining the distance change rate based on the ratio of the difference in spherical distance between adjacent time points to the time difference; calculating the heading deviation between the target vessel's current heading and the ideal heading towards the candidate port; calculating the speed change rate based on the speed data at consecutive time points; and fusing the distance change rate, heading deviation, and speed change rate to obtain the normalized dynamic approach degree.

[0029] In this technical solution, the semi-versus formula is a mathematical formula used to calculate the spherical distance between two points on the Earth's surface. It is applicable to calculating the true distance between a ship and a candidate port based on latitude and longitude coordinates and can eliminate calculation errors caused by the curvature of the Earth.

[0030] The spherical distance at consecutive time points refers to multiple consecutive distance values ​​between the ship's current position and the candidate port coordinates at adjacent consecutive time stamps, calculated using the semi-sine formula. It is used to reflect the changes in the distance between the ship and the candidate port.

[0031] The rate of change of distance is the ratio of the difference in spherical distance between adjacent time points to the time difference. It is used to characterize the speed at which a ship approaches or moves away from a candidate port. A positive value indicates that the ship is approaching the candidate port, while a negative value indicates that the ship is moving away from the candidate port.

[0032] The ideal course refers to the optimal sailing direction from the ship's current position to the candidate port, which is the heading angle corresponding to the line connecting the ship's current position and the coordinates of the candidate port.

[0033] Heading deviation refers to the angle between a ship's current actual heading and its ideal heading toward a candidate port. It is used to characterize the consistency between the ship's sailing direction and the target port. The smaller the angle, the more accurate the heading.

[0034] The rate of change of speed is the ratio of the difference in speed between adjacent time points to the difference in time. It is used to characterize the trend of change in ship speed. A negative value indicates that the ship is decelerating, and a positive value indicates that the ship is accelerating. Ships usually show a deceleration trend before berthing at port.

[0035] Normalization refers to converting indicators with different dimensions, such as the rate of change of distance, heading deviation, and rate of change of speed, into dimensionless data within the range of [0,1], so that the indicators are comparable and can be easily integrated for calculation.

[0036] This application calculates spherical distance using the semi-sine formula, and combines the fusion calculation and normalization of distance change rate, heading deviation, and speed change rate to obtain a dynamic approach degree that comprehensively reflects the degree to which a ship approaches a candidate port. This achieves a quantitative characterization of ship approach behavior and provides accurate indicator support for subsequent convergence analysis and destination port inference.

[0037] First, based on the ship's latitude and longitude and the candidate port coordinates, the spherical distance at consecutive time points is calculated using the semi-versus formula. This accurately reflects the true geographical distance between the ship and the candidate port, avoiding distance calculation errors caused by the Earth's curvature. Next, the rate of change of distance is calculated by the ratio of the difference in spherical distance between adjacent time points to the time difference, precisely capturing the speed trend of the ship approaching or moving away from the candidate port. Then, the course deviation between the ship's current course and its ideal course is calculated, clarifying the consistency between the ship's sailing direction and the target port. Simultaneously, the rate of change of speed is calculated based on the speed data at consecutive time points, capturing the ship's acceleration or deceleration trends, especially deceleration behavior before berthing. Finally, by setting reasonable normalization weight coefficients, the rate of change of distance, course deviation, and rate of change of speed are integrated and normalized to the [0,1] range to obtain the dynamic approach degree. This index integrates core dynamic factors such as the change in distance between the ship and the port, course consistency, and speed adjustment, enabling a comprehensive and accurate quantification of the degree to which the ship approaches the candidate port. For example, when a cargo ship is sailing toward a candidate port, the distance change rate is positive and gradually increases (the speed increases as it approaches), the heading deviation remains within 10° (the heading is accurate), and the speed change rate is negative and the absolute value gradually increases (the speed continues to decrease). The dynamic convergence obtained through fusion calculation continues to increase, clearly reflecting the trend of the cargo ship berthing toward the candidate port, and providing a reliable quantitative basis for subsequent convergence analysis.

[0038] In some technical solutions of this application, convergence analysis is performed on multiple dynamic convergence degrees to determine candidate ports that meet the convergence conditions from multiple port data as target ports for the target vessel. This includes analyzing the mean change trend and volatility of the dynamic convergence degree sequences corresponding to each candidate port. When the mean dynamic convergence degree corresponding to any candidate port continuously increases and the volatility continuously decreases and tends to stabilize, the candidate port is determined to meet the convergence conditions. Candidate ports that meet the convergence conditions and whose convergence characteristics meet preset conditions are determined as target ports for the target vessel.

[0039] In this technical solution, the dynamic approach sequence refers to a set of continuous dynamic approach values ​​arranged in chronological order for a candidate port, which can reflect the changing trend of the approach of ships relative to the candidate port over time.

[0040] The mean change trend refers to the direction of change of the average value of the dynamic approach sequence over time, including three situations: continuous increase, continuous decrease, or remaining stable. A continuous increase in the mean indicates that the approach trend of ships is strengthening.

[0041] Volatility refers to the degree of dispersion of a dynamic convergence sequence, usually expressed as variance or standard deviation. The smaller the volatility, the more stable the change in dynamic convergence.

[0042] The convergence feature is considered optimal if it meets the preset conditions. In other words, among multiple candidate ports that meet the convergence conditions, the candidate port with the fastest growth rate of the mean dynamic approach degree, the most significant decrease in volatility, and the lowest stable level is the port with the most significant ship approach trend.

[0043] This application analyzes the mean change trend and volatility of the dynamic convergence sequence corresponding to each candidate port to accurately select the candidate port that meets the convergence condition and has the best convergence characteristics as the real target port. This enables automatic inference of the real berthing port of a ship under the condition of unknown destination port, effectively solving the problem of destination port identification in densely populated areas of multiple ports and complex navigation scenarios.

[0044] For multiple candidate ports, each port corresponds to a set of dynamic approach degree sequences that change over time. A single point in time's dynamic approach degree value is insufficient to accurately determine a ship's true berthing intention; therefore, time series analysis is needed to uncover trend characteristics. First, the mean change trend and volatility of each dynamic approach degree sequence are analyzed. When the mean dynamic approach degree for a candidate port continuously increases, it indicates that the ship's approach to that port is constantly strengthening. When the volatility continuously decreases and tends to stabilize, it indicates that the ship's approach behavior towards that port is stable and not caused by accidental adjustments in course or speed. At this point, the candidate port is determined to meet the convergence condition, meaning the ship has a high probability of berthing at that port. Subsequently, among the candidate ports that meet the convergence condition, the port with the best convergence characteristics is selected as the true target port, ensuring the uniqueness and accuracy of the destination port inference. For example, in the densely populated port area of ​​the Yangtze River Delta, three candidate ports exist near the final stage of a ship's trajectory. Analysis of the dynamic convergence sequences corresponding to each port reveals that the mean dynamic convergence of port A continuously increases from 0.3 to 0.8, while its volatility decreases from 0.15 to 0.03 and tends to stabilize. The mean dynamic convergence of port B fluctuates between 0.4 and 0.5, without a significant upward trend. Although the mean dynamic convergence of port C increases, its volatility remains consistently above 0.1. Therefore, port A is determined to meet the convergence criteria and has the optimal convergence characteristic, thus identifying it as the ship's true destination port. This successfully achieves automatic destination port inference in complex environments.

[0045] In some technical solutions of this application, derived features include the mean, volatility, trend coefficient, speed change rate, heading stability, and distance change rate of dynamic approach, as well as the target vessel's ship type, length, and gross tonnage data. The trend coefficient of dynamic approach is the mean rate of change of dynamic approach within a continuous time window. Heading stability is the standard deviation of heading deviation within a continuous time window.

[0046] In this technical solution, the mean refers to the average value of the dynamic approach sequence within a continuous time window, reflecting the overall level of the ship's approach to the target port. Volatility refers to the standard deviation of the dynamic approach sequence within a continuous time window, reflecting the stability of the ship's approach trend. The trend coefficient refers to the average rate of change of the dynamic approach sequence within a continuous time window, quantifying the strength of the upward or downward trend in dynamic approach. Heading stability refers to the standard deviation of the heading deviation within a continuous time window, reflecting the stability of the ship's heading; a smaller standard deviation indicates greater heading stability. Ship static characteristics refer to the inherent attributes of a ship that do not change during navigation, including ship type (cargo ship, tanker, passenger ship, etc.), length, and gross tonnage. These characteristics affect the ship's berthing behavior and navigation characteristics.

[0047] This application enriches the calculation dimensions of arrival confidence by clarifying the specific types and calculation methods of derived features, and combines the trend characteristics of dynamic approach, the stability characteristics of ship navigation, and the inherent attributes of ships to improve the comprehensiveness and accuracy of arrival confidence calculation, providing a more reliable quantitative basis for the accurate determination of arrival status.

[0048] The calculation of arrival confidence requires full consideration of various key factors affecting ship arrival; relying solely on the dynamic approach rate as a single indicator is insufficient to comprehensively reflect the likelihood of a ship's arrival. By incorporating the mean, volatility, and trend coefficient of the dynamic approach rate into derived features, the approach behavior of ships can be characterized from three dimensions: overall level, stability, and trend. Incorporating the rate of change of speed, heading stability, and distance change rate into derived features allows for the capture of dynamic adjustments in the ship's navigation status. Incorporating ship type, length, and gross tonnage data into derived features allows for adaptation to the differences in berthing characteristics of different ships (e.g., large oil tankers have longer deceleration distances and longer berthing preparation times). Specifically, the trend coefficient of the dynamic approach rate is calculated using the mean of the rate of change of the dynamic approach rate within a continuous time window, accurately quantifying the strength of the approach trend. Heading stability is calculated using the standard deviation of the heading deviation within a continuous time window, accurately reflecting the stability of the ship's heading. These derived features, combined with dynamic approach scores, form a multi-dimensional and comprehensive feature system. The arrival confidence scores calculated after inputting these features into the model can more comprehensively and accurately characterize the likelihood of a ship's arrival. For example, when a large cargo ship and a small passenger ship sail towards the same port, both have a mean dynamic approach score of 0.7. However, the large cargo ship has a trend coefficient of 0.05 (a gentle approach trend) and a course stability of 2° (stable course), while the small passenger ship has a trend coefficient of 0.1 (a strong approach trend) and a course stability of 5° (significant course fluctuation). Combined with static features such as ship type and length, the model calculates arrival confidence scores of 0.72 and 0.68 respectively, accurately reflecting the differences in arrival probability between different ships and providing precise support for subsequent status determination.

[0049] In some technical solutions of this application, the arrival status of a target vessel is determined based on the arrival confidence level, including: when the arrival confidence level is greater than or equal to a first threshold, the target vessel is determined to be in an arrived state; when the arrival confidence level is greater than or equal to a second threshold but less than the first threshold, and there is a continuous upward trend, the target vessel is determined to be in an expected arrival state; when the arrival confidence level is less than the second threshold, the target vessel is determined to be in an unarrived state. Wherein, the first threshold is greater than the second threshold, the first threshold is 0.8, and the second threshold is 0.6.

[0050] In this technical solution, the first threshold, set at 0.8, is the critical confidence level for determining whether a target vessel has arrived at the port. This threshold is validated based on extensive historical berthing data, ensuring high reliability in determining the arrival status. The second threshold, set at 0.6, is the critical confidence level for distinguishing between expected arrival and non-arrival status. A value below this threshold indicates a low probability of vessel arrival. "Arrived" refers to the target vessel having completed berthing and is stably docked at the target port, corresponding to an arrival confidence level greater than or equal to the first threshold. "Expected" refers to the target vessel continuously approaching the target port, with an increasing probability of arrival, but not yet having completed berthing, corresponding to an arrival confidence level between the second and first thresholds and showing a continuous upward trend. "Non-arrival" refers to the target vessel not approaching any port or having an extremely low probability of arrival, corresponding to an arrival confidence level less than the second threshold. A continuous upward trend indicates that the arrival confidence level gradually increases over multiple consecutive time windows, suggesting a continuously strengthening probability of vessel arrival.

[0051] This application achieves refined classification of arrival status by setting a first threshold and a second threshold, combined with the magnitude and trend of arrival confidence scores. This allows for accurate identification of already arrived status, prediction of expected arrival status, and effective differentiation of non-arrival status, meeting the decision-making needs of different scenarios in port scheduling and improving the flexibility and practicality of arrival status determination. The first threshold is set at 0.8, a high-confidence threshold verified based on a large amount of historical ship berthing data. When the arrival confidence score is greater than or equal to 0.8, it indicates a very high probability of ship arrival, and this status is stable, accurately determining it as already arrived, avoiding the randomness that may exist in high confidence scores from a single sample. The second threshold is set at 0.6. Below this threshold, the probability of ship arrival is low, and it is determined as non-arrival status, effectively filtering out interference from non-arriving ships and reducing false alarms. When the arrival confidence level is between 0.6 and 0.8 and shows a continuous upward trend, it indicates that the vessel is gradually approaching the target port, and the probability of arrival is constantly increasing. This is determined as an expected arrival status, providing port scheduling with advance preparation time and solving the deficiency of traditional methods that cannot predict in advance. For example, when an oil tanker is sailing towards the target port, its arrival confidence level gradually increases from 0.55 to 0.75, and shows a continuous upward trend for five consecutive time windows. The model determines this as an expected arrival status, and the port scheduling department can arrange loading and unloading equipment, personnel, and berths in advance, improving operational efficiency. When the oil tanker continues to approach the port, and the arrival confidence level rises to 0.82 and remains stable for three time windows, the model determines this as an already arrived status, triggering port operation procedures. On the other hand, for a fishing vessel operating only in the waters near the port, its arrival confidence level is consistently below 0.6. The model determines this as an unarrived status, avoiding unnecessary interference with port scheduling and achieving precise and refined management of arrival status.

[0052] A second aspect of the present invention provides a vessel arrival status determination device, comprising: a first acquisition module, a first determination module, a second determination module, a third determination module, a fourth determination module, and a fifth determination module. The first acquisition module acquires navigation dynamic data of a target vessel and determines the final position data of the target vessel's trajectory based on the navigation dynamic data. The first determination module determines multiple candidate ports within a preset navigation range based on the final position data of the trajectory and acquires port data for each candidate port. The second determination module determines the dynamic proximity of the target vessel relative to each candidate port based on the multiple port data. The third determination module performs convergence analysis on the multiple dynamic proximity values ​​and determines the candidate ports among the multiple port data that meet the convergence conditions as the target ports for the target vessel to dock. The fourth determination module determines the arrival confidence level based on the dynamic proximity and derived features corresponding to the target port. The fifth determination module determines the arrival status of the target vessel based on the arrival confidence level.

[0053] The vessel arrival status determination device in this application constructs a complete technical chain of "data acquisition - candidate port generation - dynamic proximity calculation - destination port inference - confidence assessment - status determination" through the collaborative work of the first acquisition module, the first determination module, the second determination module, the third determination module, the fourth determination module, and the fifth determination module. The core technical effect is to break through the dependence on the vessel's self-reported destination port or preset port coordinates, and achieve autonomous inference of unknown destination ports by automatically generating candidate ports and analyzing the convergence characteristics of dynamic proximity. At the same time, it integrates dynamic proximity and derived features to calculate the arrival confidence, accurately determine the three states of arrival, expected arrival, and not yet arrived, effectively solve the false alarm problem in densely populated areas with multiple ports and complex navigation scenarios, shorten the arrival identification delay, improve the device's adaptability to different sea areas and port environments, and enhance the coverage and accuracy of arrival status identification, providing real-time and reliable decision support for port scheduling and shipping management.

[0054] A third aspect of the present invention provides a vessel arrival status determination apparatus, comprising a processor and a memory, wherein the memory stores a program or instructions, and the processor, when executing the program or instructions in the memory, implements the steps of the vessel arrival status determination method as described in any of the above-described technical solutions. Therefore, the vessel arrival status determination apparatus possesses all the beneficial effects of the vessel arrival status determination method as described in any of the above-described technical solutions.

[0055] A fourth aspect of the present invention provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of the ship arrival status determination method as described in any of the above-described technical solutions. Therefore, the readable storage medium possesses all the beneficial effects of the ship arrival status determination method as described in any of the above-described technical solutions.

[0056] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0057] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0058] Figure 1 This is one of the flowcharts illustrating a method for determining the arrival status of a ship according to an embodiment of the present invention;

[0059] Figure 2 This is a second schematic flowchart of a method for determining the arrival status of a ship according to an embodiment of the present invention;

[0060] Figure 3 This is a structural diagram of a potential port candidate generation module according to an embodiment of the present invention;

[0061] Figure 4 This is a schematic diagram of dynamic convergence analysis according to an embodiment of the present invention;

[0062] Figure 5 This is a flowchart illustrating the destination port determination and arrival port assessment according to an embodiment of the present invention.

[0063] Figure 6 This is a schematic diagram of the arrival confidence curve according to an embodiment of the present invention;

[0064] Figure 7 This is one of the schematic block diagrams of a ship arrival status determination device according to an embodiment of the present invention;

[0065] Figure 8 This is a second schematic block diagram of a ship arrival status determination device according to an embodiment of the present invention. Detailed Implementation

[0066] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0067] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0068] The following reference Figures 1 to 8A method, apparatus, and readable storage medium for determining the arrival status of a ship are described according to some embodiments of the present invention.

[0069] like Figure 1 As shown in the embodiments of this application, a method for determining the arrival status of a ship is provided, the steps of which include:

[0070] Step 102: Obtain the navigation dynamic data of the target vessel, and determine the final position data of the target vessel's track based on the navigation dynamic data.

[0071] Step 104: Based on the terminal position data of the flight path, determine multiple candidate ports within the preset navigation range and obtain port data for each candidate port.

[0072] Step 106: Determine the dynamic proximity of the target vessel to each candidate port based on data from multiple ports.

[0073] Step 108: Perform convergence analysis on multiple dynamic convergence values ​​to identify candidate ports that meet the convergence criteria from multiple port data as the target ports for the target vessels to dock.

[0074] Step 110: Determine the arrival confidence level based on the dynamic proximity and derived characteristics of the target port.

[0075] Step 112: Determine the arrival status of the target vessel based on the arrival confidence level.

[0076] The navigation dynamic data provided in this application refers to real-time navigation-related data collected by the target vessel through the AIS system, including latitude and longitude, speed, heading, timestamp, etc., which is the core data support reflecting the real-time motion status of the vessel. The track end position data refers to the set of key position information of the target vessel in the later stage of navigation, which is extracted from the navigation dynamic data after preprocessing, and is used to clarify the search benchmark for potential port candidates.

[0077] The preset navigation range refers to a geographical search area centered on the end position of the vessel's trajectory, within 50 nautical miles, used to filter potential ports of call. Candidate ports refer to all ports within the preset navigation range, selected through geographical matching to identify potential berthing targets. Port data refers to the core attribute information of each candidate port, including coordinates, name, and boundaries, providing basic coordinates and features for dynamic proximity calculations.

[0078] Dynamic approach degree refers to a normalized index that characterizes the degree of approach of a target ship to a candidate port. The value range is [0,1]. It is calculated by integrating the distance change rate, heading deviation and speed change rate. The larger the value, the more obvious the approach trend.

[0079] Convergence analysis refers to the time series analysis of the dynamic convergence sequences corresponding to multiple candidate ports, and the analysis process of judging whether ships are really approaching a certain port by the trend of mean change and volatility.

[0080] The convergence condition refers to the core criteria for determining whether a candidate port is the actual target port, namely, the mean of dynamic convergence continuously increases and the volatility continuously decreases and tends to stabilize.

[0081] The target port refers to the port that the ship actually berths at, selected from multiple candidate ports, meets the convergence criteria, and has the best convergence characteristics.

[0082] Derived features refer to statistical features and inherent ship attributes derived from dynamic approach and navigation dynamic data, including the mean, volatility, and trend coefficient of dynamic approach, which are used to enrich the calculation dimensions of arrival confidence.

[0083] Arrival confidence level refers to the probability value that represents the possibility of a target vessel arriving at the port. The value range is [0,1]. It is calculated by inputting the dynamic approach degree and derived features into the trained model, providing a quantitative basis for determining the arrival status.

[0084] Arrival status refers to the classification of the target vessel's navigation status, including three categories: arrived, expected to arrive, and not yet arrived, which are determined based on arrival confidence and preset thresholds.

[0085] Based on the above data and steps, the solution proposed in this application addresses the technical problems of relying on ships' self-reported destination ports or preset port coordinates for determining ship arrival. When destination port information is missing, incorrect, or the berthing target in a densely populated area with multiple ports is unclear, the solution cannot effectively identify the ship's true destination port and arrival status, leading to high false alarm rates, poor adaptability, and a lack of advance prediction capabilities. The solution achieves automatic inference of the ship's true destination port and accurate, advance determination of its arrival status under unknown destination port conditions, significantly improving port scheduling efficiency and the level of intelligent shipping management. Specifically, by acquiring the target ship's navigation dynamic data and determining the final position data of its trajectory, it eliminates the need to rely on the ship's self-reported destination port information, locking in potential berthing areas solely based on the ship's actual navigation trajectory, thus overcoming the limitation of dependence on declared destination port information. Next, using the final position data of the trajectory as a benchmark, multiple candidate ports are screened within a preset navigation range, and port data is acquired to construct a comprehensive set of potential berthing targets, avoiding omissions or misjudgments caused by judging a single port. Then, a dynamic approach score is calculated for each candidate port, comprehensively quantifying dynamic behaviors such as changes in distance between the ship and the port, course consistency, and speed adjustments. This allows for accurate capture of the trend of ships approaching a specific port. Subsequently, through convergence analysis of multiple dynamic approach scores, candidate ports with continuously increasing mean and stable volatility are selected as the true target ports. This enables autonomous inference of the destination port from a behavioral pattern perspective, effectively distinguishing between transit and arrival behaviors and solving the challenge of destination port identification in complex navigation scenarios. Next, the arrival confidence score is calculated based on the dynamic approach score and derived features corresponding to the target port, integrating dynamic trends and inherent ship attributes to improve the comprehensiveness and accuracy of arrival probability assessment. Finally, the arrival status is determined based on the arrival confidence score and a preset threshold. This not only clearly identifies the arrival and non-arrival states but also predicts the expected arrival status in advance, providing sufficient preparation time for port scheduling, significantly shortening the determination delay, and overcoming the lag in static rule-based determinations. The entire process forms a complete technical chain of trajectory analysis, candidate generation, proximity calculation, destination port inference, confidence assessment, and state determination, realizing the improvement from relying on the input destination to autonomously judging the destination, and significantly improving the coverage and accuracy of ship arrival identification in the scenario of unknown destination port.

[0086] In some embodiments of this application, acquiring the navigation dynamic data of the target vessel and determining the final position data of the target vessel's track based on the navigation dynamic data includes: acquiring the navigation dynamic data of the target vessel, removing outliers and drift data from the navigation dynamic data, and organizing the data into track data in chronological order. The final position data of the target vessel's track is then determined based on the track data.

[0087] In this embodiment, outliers refer to extreme data in navigation dynamics that deviate from normal navigation patterns, such as latitude and longitude jumps or abnormal speed spikes caused by equipment failure or signal interference.

[0088] Drift data refers to offset data in navigation dynamics that does not match the actual navigation trajectory of a ship, such as position drift data caused by AIS signal delay.

[0089] Track data refers to a continuous data sequence that accurately reflects a ship's navigation path, after outliers and drift data have been removed and arranged in chronological order.

[0090] This application generates real and continuous track data by removing outliers and drift data from navigation dynamic data and organizing the data in chronological order. This provides a reliable data foundation for the accurate determination of the final position data of the track and avoids the subsequent candidate port selection bias and destination port inference errors caused by noise in the original data.

[0091] Navigational dynamic data is susceptible to factors such as equipment failure, signal interference, and transmission delays during acquisition, resulting in outliers and drift data. Directly using this data for analysis can distort the trajectory, thus affecting the accuracy of the trajectory's final position data. By acquiring the target vessel's navigational dynamic data and first removing outliers and drift data, data noise can be filtered out, restoring the vessel's true navigation trajectory. The remaining valid data is then processed chronologically to form continuous trajectory data, ensuring temporal consistency and guaranteeing accurate extraction of the trajectory's final position data. For example, if a vessel experiences a temporary AIS equipment malfunction during navigation, generating a set of abnormal data with latitude and longitude jumps far from the route, the processing steps in this embodiment can remove this abnormal data. The remaining data, processed chronologically, forms trajectory data that accurately reflects the vessel's navigation path. The trajectory's final position data determined based on this data can accurately pinpoint key areas in the later stages of the vessel's navigation, ensuring the effectiveness of subsequent candidate port selection.

[0092] In some embodiments of this application, determining the final position data of the target vessel's track based on track data includes: smoothing the speed and heading data in the track data, standardizing the coordinate, time, and distance units of all data in the track data, and generating standardized track data. The final position data of the target vessel's track is then determined based on the standardized track data.

[0093] In this embodiment, smoothing refers to filtering the speed and heading data in the track data to eliminate high-frequency fluctuations, making the data changes smoother and better reflecting the true trend of ship navigation. Coordinate units refer to units of measurement representing latitude and longitude, such as degrees, minutes, and seconds, and must be standardized to a consistent unit format to ensure the accuracy of distance calculations. Time units refer to units of measurement representing time, such as seconds and minutes, and must be standardized to a consistent unit format to ensure the consistency of rate of change calculations. Distance units refer to units of measurement representing distance, such as meters and nautical miles, and must be standardized to a consistent unit format to ensure the consistency of distance-related indicator calculations. Standardized track data refers to track data after smoothing and unit standardization, possessing the characteristics of stable data, unified format, and reliable accuracy, and is the core basis for accurately determining the final position data of the track segment.

[0094] This application generates standardized track data by smoothing and unifying the units of speed and heading in track data. This further improves the reliability and consistency of track data, ensuring the accuracy of the calculation of the final position data of the track, and laying a high-precision data foundation for subsequent candidate port generation and dynamic approach calculation. Speed ​​and heading data in track data may exhibit high-frequency fluctuations, such as data jitter caused by slight adjustments to the ship's heading or speed. These fluctuations can affect the judgment of the ship's true navigation trend. Simultaneously, the original dynamic navigation data may have inconsistent units for coordinates, time, and distance, directly affecting the accuracy of subsequent calculations. By smoothing the speed and heading data in track data, meaningless high-frequency fluctuations can be filtered out, highlighting the overall trend of the ship's navigation. By unifying the units for coordinates, time, and distance, all data formats are ensured to be consistent, avoiding calculation errors caused by unit differences. For example, a ship's speed data may fluctuate frequently due to wind and waves, and its heading data may experience slight jitter due to manual fine-tuning. After smoothing, the trends in speed and heading data are more stable, accurately reflecting the ship's navigation intentions. Meanwhile, the distance data in the original data that was in meters and distance data that was in nautical miles were unified into nautical miles to ensure the accuracy of subsequent calculations of spherical distance and distance change rate. The accuracy of the final position data of the track determined based on the standardized track data was significantly improved, which provided strong support for the accurate selection of candidate ports.

[0095] In some embodiments of this application, the dynamic approach degree of the target vessel relative to each candidate port is determined based on multiple port data, including: calculating the spherical distance at consecutive time points using the semi-versus formula based on the vessel's latitude and longitude and the candidate port coordinates; determining the distance change rate based on the ratio of the difference in spherical distance between adjacent time points to the time difference; calculating the heading deviation between the target vessel's current heading and the ideal heading towards the candidate port; calculating the speed change rate based on the speed data at consecutive time points; and fusing the distance change rate, heading deviation, and speed change rate to obtain the normalized dynamic approach degree.

[0096] In this embodiment, the semi-versus formula is a mathematical formula used to calculate the spherical distance between two points on the Earth's surface. It is applicable to calculating the true distance between a ship and a candidate port based on latitude and longitude coordinates and can eliminate calculation errors caused by the curvature of the Earth.

[0097] The spherical distance at consecutive time points refers to multiple consecutive distance values ​​between the ship's current position and the candidate port coordinates at adjacent consecutive time stamps, calculated using the semi-sine formula. It is used to reflect the changes in the distance between the ship and the candidate port.

[0098] The rate of change of distance is the ratio of the difference in spherical distance between adjacent time points to the time difference. It is used to characterize the speed at which a ship approaches or moves away from a candidate port. A positive value indicates that the ship is approaching the candidate port, while a negative value indicates that the ship is moving away from the candidate port.

[0099] The ideal course refers to the optimal sailing direction from the ship's current position to the candidate port, which is the heading angle corresponding to the line connecting the ship's current position and the coordinates of the candidate port.

[0100] Heading deviation refers to the angle between a ship's current actual heading and its ideal heading toward a candidate port. It is used to characterize the consistency between the ship's sailing direction and the target port. The smaller the angle, the more accurate the heading.

[0101] The rate of change of speed is the ratio of the difference in speed between adjacent time points to the difference in time. It is used to characterize the trend of change in ship speed. A negative value indicates that the ship is decelerating, and a positive value indicates that the ship is accelerating. Ships usually show a deceleration trend before berthing at port.

[0102] Normalization refers to converting indicators with different dimensions, such as the rate of change of distance, heading deviation, and rate of change of speed, into dimensionless data within the range of [0,1], so that the indicators are comparable and can be easily integrated for calculation.

[0103] This application calculates spherical distance using the semi-sine formula, and combines the fusion calculation and normalization of distance change rate, heading deviation, and speed change rate to obtain a dynamic approach degree that comprehensively reflects the degree to which a ship approaches a candidate port. This achieves a quantitative characterization of ship approach behavior and provides accurate indicator support for subsequent convergence analysis and destination port inference.

[0104] First, based on the ship's latitude and longitude and the candidate port coordinates, the spherical distance at consecutive time points is calculated using the semi-versus formula. This accurately reflects the true geographical distance between the ship and the candidate port, avoiding distance calculation errors caused by the Earth's curvature. Next, the rate of change of distance is calculated by the ratio of the difference in spherical distance between adjacent time points to the time difference, precisely capturing the speed trend of the ship approaching or moving away from the candidate port. Then, the course deviation between the ship's current course and its ideal course is calculated, clarifying the consistency between the ship's sailing direction and the target port. Simultaneously, the rate of change of speed is calculated based on the speed data at consecutive time points, capturing the ship's acceleration or deceleration trends, especially deceleration behavior before berthing. Finally, by setting reasonable normalization weight coefficients, the rate of change of distance, course deviation, and rate of change of speed are integrated and normalized to the [0,1] range to obtain the dynamic approach degree. This index integrates core dynamic factors such as the change in distance between the ship and the port, course consistency, and speed adjustment, enabling a comprehensive and accurate quantification of the degree to which the ship approaches the candidate port. For example, when a cargo ship is sailing toward a candidate port, the distance change rate is positive and gradually increases (the speed increases as it approaches), the heading deviation remains within 10° (the heading is accurate), and the speed change rate is negative and the absolute value gradually increases (the speed continues to decrease). The dynamic convergence obtained through fusion calculation continues to increase, clearly reflecting the trend of the cargo ship berthing toward the candidate port, and providing a reliable quantitative basis for subsequent convergence analysis.

[0105] In some embodiments of this application, convergence analysis is performed on multiple dynamic convergence degrees to determine candidate ports that meet the convergence criteria from multiple port data as target ports for the target vessel. This includes analyzing the mean change trend and volatility of the dynamic convergence degree sequences corresponding to each candidate port. When the mean dynamic convergence degree corresponding to any candidate port continuously increases and the volatility continuously decreases and tends to stabilize, the candidate port is determined to meet the convergence criteria. Candidate ports that meet the convergence criteria and whose convergence characteristics meet preset conditions are determined as target ports for the target vessel.

[0106] In this embodiment, the dynamic approach sequence refers to a set of continuous dynamic approach values ​​arranged in chronological order for a candidate port, which can reflect the changing trend of the approach of the ship relative to the candidate port over time.

[0107] The mean change trend refers to the direction of change of the average value of the dynamic approach sequence over time, including three situations: continuous increase, continuous decrease, or remaining stable. A continuous increase in the mean indicates that the approach trend of ships is strengthening.

[0108] Volatility refers to the degree of dispersion of a dynamic convergence sequence, usually expressed as variance or standard deviation. The smaller the volatility, the more stable the change in dynamic convergence.

[0109] The convergence feature is considered optimal if it meets the preset conditions. In other words, among multiple candidate ports that meet the convergence conditions, the candidate port with the fastest growth rate of the mean dynamic approach degree, the most significant decrease in volatility, and the lowest stable level is the port with the most significant ship approach trend.

[0110] This application analyzes the mean change trend and volatility of the dynamic convergence sequence corresponding to each candidate port to accurately select the candidate port that meets the convergence condition and has the best convergence characteristics as the real target port. This enables automatic inference of the real berthing port of a ship under the condition of unknown destination port, effectively solving the problem of destination port identification in densely populated areas of multiple ports and complex navigation scenarios.

[0111] For multiple candidate ports, each port corresponds to a set of dynamic approach degree sequences that change over time. A single point in time's dynamic approach degree value is insufficient to accurately determine a ship's true berthing intention; therefore, time series analysis is needed to uncover trend characteristics. First, the mean change trend and volatility of each dynamic approach degree sequence are analyzed. When the mean dynamic approach degree for a candidate port continuously increases, it indicates that the ship's approach to that port is constantly strengthening. When the volatility continuously decreases and tends to stabilize, it indicates that the ship's approach behavior towards that port is stable and not caused by accidental adjustments in course or speed. At this point, the candidate port is determined to meet the convergence condition, meaning the ship has a high probability of berthing at that port. Subsequently, among the candidate ports that meet the convergence condition, the port with the best convergence characteristics is selected as the true target port, ensuring the uniqueness and accuracy of the destination port inference. For example, in the densely populated port area of ​​the Yangtze River Delta, three candidate ports exist near the final stage of a ship's trajectory. Analysis of the dynamic convergence sequences corresponding to each port reveals that the mean dynamic convergence of port A continuously increases from 0.3 to 0.8, while its volatility decreases from 0.15 to 0.03 and tends to stabilize. The mean dynamic convergence of port B fluctuates between 0.4 and 0.5, without a significant upward trend. Although the mean dynamic convergence of port C increases, its volatility remains consistently above 0.1. Therefore, port A is determined to meet the convergence criteria and has the optimal convergence characteristic, thus identifying it as the ship's true destination port. This successfully achieves automatic destination port inference in complex environments.

[0112] In some embodiments of this application, derived features include the mean, volatility, trend coefficient, rate of change of speed, heading stability, and rate of change of distance of the dynamic approach degree, as well as the target vessel's ship type, length, and gross tonnage data. The trend coefficient of the dynamic approach degree is the mean rate of change of the dynamic approach degree within a continuous time window. The heading stability is the standard deviation of the heading deviation within a continuous time window.

[0113] In this embodiment, the mean refers to the average value of the dynamic approach sequence within a continuous time window, reflecting the overall level of the ship's approach to the target port. Volatility refers to the standard deviation of the dynamic approach sequence within a continuous time window, reflecting the stability of the ship's approach trend. The trend coefficient refers to the average rate of change of the dynamic approach sequence within a continuous time window, quantifying the strength of the upward or downward trend in dynamic approach. Heading stability refers to the standard deviation of the heading deviation within a continuous time window, reflecting the stability of the ship's heading; a smaller standard deviation indicates greater heading stability. Ship static characteristics refer to the inherent attributes of a ship that do not change during navigation, including ship type (cargo ship, tanker, passenger ship, etc.), length, and gross tonnage. These characteristics affect the ship's berthing behavior and navigation characteristics.

[0114] This application enriches the calculation dimensions of arrival confidence by clarifying the specific types and calculation methods of derived features, and combines the trend characteristics of dynamic approach, the stability characteristics of ship navigation, and the inherent attributes of ships to improve the comprehensiveness and accuracy of arrival confidence calculation, providing a more reliable quantitative basis for the accurate determination of arrival status.

[0115] The calculation of arrival confidence requires full consideration of various key factors affecting ship arrival; relying solely on the dynamic approach rate as a single indicator is insufficient to comprehensively reflect the likelihood of a ship's arrival. By incorporating the mean, volatility, and trend coefficient of the dynamic approach rate into derived features, the approach behavior of ships can be characterized from three dimensions: overall level, stability, and trend. Incorporating the rate of change of speed, heading stability, and distance change rate into derived features allows for the capture of dynamic adjustments in the ship's navigation status. Incorporating ship type, length, and gross tonnage data into derived features allows for adaptation to the differences in berthing characteristics of different ships (e.g., large oil tankers have longer deceleration distances and longer berthing preparation times). Specifically, the trend coefficient of the dynamic approach rate is calculated using the mean of the rate of change of the dynamic approach rate within a continuous time window, accurately quantifying the strength of the approach trend. Heading stability is calculated using the standard deviation of the heading deviation within a continuous time window, accurately reflecting the stability of the ship's heading. These derived features, combined with dynamic approach scores, form a multi-dimensional and comprehensive feature system. The arrival confidence scores calculated after inputting these features into the model can more comprehensively and accurately characterize the likelihood of a ship's arrival. For example, when a large cargo ship and a small passenger ship sail towards the same port, both have a mean dynamic approach score of 0.7. However, the large cargo ship has a trend coefficient of 0.05 (a gentle approach trend) and a course stability of 2° (stable course), while the small passenger ship has a trend coefficient of 0.1 (a strong approach trend) and a course stability of 5° (significant course fluctuation). Combined with static features such as ship type and length, the model calculates arrival confidence scores of 0.72 and 0.68 respectively, accurately reflecting the differences in arrival probability between different ships and providing precise support for subsequent status determination.

[0116] In some embodiments of this application, determining the arrival status of a target vessel based on arrival confidence includes: determining that the target vessel is in an arrived state when the arrival confidence is greater than or equal to a first threshold; determining that the target vessel is in an expected arrival state when the arrival confidence is greater than or equal to a second threshold but less than the first threshold and shows a continuous upward trend; and determining that the target vessel is in an unarrived state when the arrival confidence is less than the second threshold. Wherein, the first threshold is greater than the second threshold, the first threshold is 0.8, and the second threshold is 0.6.

[0117] In this embodiment, the first threshold is a critical value of 0.8 for determining the target vessel's arrival confidence level. This threshold is verified based on a large amount of historical berthing data, ensuring high reliability in determining the arrival status. The second threshold is a critical value of 0.6 for distinguishing between expected arrival and non-arrival status. A value below this threshold indicates a low probability of the vessel arriving. "Arrived" refers to the target vessel having completed berthing and is stably docked at the target port, corresponding to an arrival confidence level greater than or equal to the first threshold. "Expected" refers to the target vessel continuously approaching the target port, with an increasing probability of arrival, but not yet having completed berthing, corresponding to an arrival confidence level between the second and first thresholds and showing a continuous upward trend. "Non-arrival" refers to the target vessel not approaching any port or having an extremely low probability of arrival, corresponding to an arrival confidence level less than the second threshold. A continuous upward trend indicates that the arrival confidence level gradually increases over multiple consecutive time windows, indicating a continuously strengthening probability of the vessel's arrival.

[0118] This application achieves refined classification of arrival status by setting a first threshold and a second threshold, combined with the magnitude and trend of arrival confidence scores. This allows for accurate identification of already arrived status, prediction of expected arrival status, and effective differentiation of non-arrival status, meeting the decision-making needs of different scenarios in port scheduling and improving the flexibility and practicality of arrival status determination. The first threshold is set at 0.8, a high-confidence threshold verified based on a large amount of historical ship berthing data. When the arrival confidence score is greater than or equal to 0.8, it indicates a very high probability of ship arrival, and this status is stable, accurately determining it as already arrived, avoiding the randomness that may exist in high confidence scores from a single sample. The second threshold is set at 0.6. Below this threshold, the probability of ship arrival is low, and it is determined as non-arrival status, effectively filtering out interference from non-arriving ships and reducing false alarms. When the arrival confidence level is between 0.6 and 0.8 and shows a continuous upward trend, it indicates that the vessel is gradually approaching the target port, and the probability of arrival is constantly increasing. This is determined as an expected arrival status, providing port scheduling with advance preparation time and solving the deficiency of traditional methods that cannot predict in advance. For example, when an oil tanker is sailing towards the target port, its arrival confidence level gradually increases from 0.55 to 0.75, and shows a continuous upward trend for five consecutive time windows. The model determines this as an expected arrival status, and the port scheduling department can arrange loading and unloading equipment, personnel, and berths in advance, improving operational efficiency. When the oil tanker continues to approach the port, and the arrival confidence level rises to 0.82 and remains stable for three time windows, the model determines this as an already arrived status, triggering port operation procedures. On the other hand, for a fishing vessel operating only in the waters near the port, its arrival confidence level is consistently below 0.6. The model determines this as an unarrived status, avoiding unnecessary interference with port scheduling and achieving precise and refined management of arrival status.

[0119] like Figure 2 As shown in the embodiments of this application, a method for determining the arrival status of a ship is provided, the steps of which include:

[0120] Step 202: Data preparation and preprocessing, collect ship AIS dynamic data (including latitude and longitude, speed, heading, timestamp, etc.), remove outliers and drift data, sort the tracks by time and perform coordinate standardization;

[0121] Step 204: Generation of potential port candidates in the region. Within 50 nautical miles of the end position of the ship's track, several potential port center points are automatically generated to form a temporary destination candidate set G={g1,g2,…,gn}.

[0122] Step 206: Dynamic approach degree calculation. For each candidate port center point gi, calculate the spherical distance Dgi(t) from the ship's current position to that point, the rate of change of distance ΔDgi / Δt, the heading deviation |Hgi|, and the rate of change of speed ΔSOG / Δt, and generate the corresponding dynamic approach degree PCIgi(t).

[0123] Step 208: Convergence analysis and port of destination determination. As navigation data accumulates, time series analysis is performed on the PCIgi(t) sequence of each candidate point. The port gk with the highest convergence characteristics (such as ΔPCI mean / Δt>0 and σPCI approaching stability) that continuously increases the PCI mean and decreases the volatility is selected and determined as the current true port of destination of the ship.

[0124] Step 210: Arrival determination. Based on the identified true destination port gk, apply the dynamic proximity determination rule (e.g., PCI ≥ 0.8 and volatility < 0.05 for a certain period of time) to determine the arrival status of the vessel. If the conditions are met, output the "Arrived" signal. If the PCI continues to rise but does not reach the threshold, output "Expected Arrival".

[0125] Step 212: Arrival identification and confidence fusion based on random forest model;

[0126] Step 214: Output and system integration. The system outputs the destination port location information, current arrival status and confidence value identified by the ship, and pushes it to the port management or monitoring system through the interface to realize real-time data interaction and predictive alerts.

[0127] like Figure 3 As shown, Figure 3 The module structure diagram for generating potential port candidates includes 5 sub-modules, specifically:

[0128] Ship AIS Time Series Data Acquisition Submodule: The core input module, responsible for real-time acquisition of dynamic AIS data of the target ship, including latitude and longitude, speed, heading, timestamp, and other information, providing raw data support for subsequent analysis and serving as the data source foundation for the entire module.

[0129] The ship historical trajectory end-segment coordinate extraction submodule preprocesses the collected AIS time-series data, organizes the trajectory in chronological order, filters out the continuous position coordinates of the ship in the later stage of navigation, forms the trajectory end-segment coordinate sequence, and clarifies the search benchmark position of potential port candidates.

[0130] Surrounding Port Matching Submodule (S22): Based on the coordinate sequence of the end segment of the track, a search range of 50 nautical miles is defined with this location as the center. All ports within this range are automatically matched, and potential ports that meet the geographical distance conditions are initially screened to form an initial candidate pool.

[0131] Candidate Port Feature Extraction Submodule: Extracts features from ports in the initial candidate pool, obtaining core attribute information for each port, including latitude and longitude location, port name, port code, and other key data, providing standardized features for the subsequent generation of candidate sets.

[0132] Candidate port output submodule: Integrates the extracted candidate port features to generate a temporary destination candidate set G={g1,g2,…,gn}, where each element gi corresponds to the complete feature data of a potential port. This set will serve as the core input for subsequent dynamic proximity calculation, completing the module's functional loop.

[0133] like Figure 4 As shown, Figure 4 Includes convergence port gk and PCI g1 PCI g2 PCI g3 Three dynamic convergence sequence curves are used, with the horizontal axis representing time points and the vertical axis representing the dynamic convergence value (0-0.9). The convergence pattern of the dynamic convergence is visually presented through these curves: the horizontal axis represents time evolution, and the vertical axis represents the dynamic convergence value. The three different curves correspond to the dynamic convergence trends of the three candidate ports. Specifically, the curve corresponding to the convergent port gk continuously rises and tends to stabilize over time, while the other two curves show no obvious upward trend or fluctuate significantly. This clearly reflects the convergence characteristic of "continuously increasing mean and continuously decreasing volatility," making the technical logic of "screening the true destination port through convergence analysis" easier to understand and intuitively demonstrating how to accurately pinpoint the actual berthing target of ships from multiple candidate ports.

[0134] like Figure 5 As shown, Figure 5 The flowchart for determining the port of destination and arrival includes the following steps:

[0135] Step 302: Input the candidate port set G={g1,g2,…,gn} and the dynamic proximity index (PCI) time series of each candidate port to provide basic data for subsequent analysis;

[0136] Step 304: Input the candidate port set and the corresponding PCI time series into the dynamic convergence analysis module, perform time series analysis on the PCI series of each candidate port, and select the optimal convergence port gk with the highest convergence characteristics that has a continuously increasing PCI mean and continuously decreasing volatility, and determine it as the actual destination port of the ship.

[0137] Step 306: Input the relevant features of the convergent optimal port gk (including the dynamic proximity sequence, derived statistical features and ship static features corresponding to gk, etc.) into the port identification model based on random forest;

[0138] Step 308: The random forest-based arrival identification model performs inference calculations on the input features and outputs the arrival confidence P_arrival (arrival confidence) and P_approach (expected arrival confidence).

[0139] Step 310: Input P_arrival and P_approach into the confidence fusion calculation module, and calculate the comprehensive confidence C_fuse=α×P_arrival+(1–α)×PCI_mean (PCI_mean is the mean of dynamic convergence) by combining the dynamic weight α (empirically taken as 0.6-0.8), thereby realizing the fusion of physical indicators and statistical model results;

[0140] Step 312: Input the comprehensive confidence level C_fuse into the rule judgment module and judge according to the preset rules: if C_fuse≥0.8, judge as "arrived at port"; if 0.6≤C_fuse<0.8, judge as "expected to arrive at port"; if C_fuse<0.6, judge as "not arrived at port".

[0141] Step 314: Integrate the rule judgment results, converge the coordinates of the optimal port gk, and the comprehensive confidence level C_fuse, and output the final result of "destination port coordinates + arrival status + confidence level value".

[0142] like Figure 6 As shown, in the model inference and confidence fusion process, the system performs smoothing filtering on the confidence value and makes a judgment based on the continuity of time, outputting the arrival status classification result. Figure 6 It contains P arrival Curve (Confidence Curve of Arrival), C fuse The curve (overall confidence curve) has time nodes on the horizontal axis (1-49) and confidence values ​​on the vertical axis (0.3-0.9), marked with "expected arrival range".

[0143] like Figure 7As shown, an embodiment of this application provides a vessel arrival status determination device 400, including: a first acquisition module 410, a first determination module 420, a second determination module 430, a third determination module 440, a fourth determination module 450, and a fifth determination module 460. The first acquisition module 410 acquires the navigation dynamic data of the target vessel and determines the final position data of the target vessel's trajectory based on the navigation dynamic data. The first determination module 420 determines multiple candidate ports within a preset navigation range based on the final position data of the trajectory and acquires port data for each candidate port. The second determination module 430 determines the dynamic proximity of the target vessel relative to each candidate port based on the multiple port data. The third determination module 440 performs convergence analysis on the multiple dynamic proximity values ​​and determines the candidate ports that meet the convergence conditions among the multiple port data as the target ports for the target vessel to dock. The fourth determination module 450 determines the arrival confidence level based on the dynamic proximity and derived features corresponding to the target port. The fifth determination module 460 determines the arrival status of the target vessel based on the arrival confidence level.

[0144] The vessel arrival status determination device 400 in this application constructs a complete technical chain of "data acquisition - candidate port generation - dynamic proximity calculation - destination port inference - confidence assessment - status determination" through the collaborative work of the first acquisition module 410, the first determination module 420, the second determination module 430, the third determination module 440, the fourth determination module 450, and the fifth determination module 460. The core technical effect is to break through the dependence of traditional devices on the vessel's self-reported destination port or preset port coordinates, and achieve autonomous inference of unknown destination ports by automatically generating candidate ports and analyzing the convergence characteristics of dynamic proximity. At the same time, it integrates dynamic proximity and derived features to calculate the arrival confidence, accurately determine the three states of arrival, expected arrival, and not yet arrived, effectively solve the false alarm problem in densely populated areas with multiple ports and complex navigation scenarios, shorten the arrival identification delay, improve the device's adaptability to different sea areas and port environments, and enhance the coverage and accuracy of arrival status identification, providing real-time and reliable decision support for port scheduling and shipping management.

[0145] like Figure 8 As shown, an embodiment of this application provides a ship arrival status determination device 500, including a processor 502 and a memory 504. The memory 504 stores programs or instructions, and when the processor 502 executes the programs or instructions in the memory 504, it implements the steps of the ship arrival status determination method as described in any of the above embodiments. Therefore, the ship arrival status determination device 500 possesses all the beneficial effects of the ship arrival status determination method as described in any of the above embodiments.

[0146] Embodiments of this application provide a readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the ship arrival status determination method as described in any of the above embodiments. Therefore, the readable storage medium possesses all the beneficial effects of the ship arrival status determination method as described in any of the above embodiments.

[0147] In the claims, description, and accompanying drawings of this invention, the term "plural" refers to two or more. Unless otherwise explicitly defined, the terms "upper," "lower," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the descriptive process, not to indicate or imply that the device or element referred to must have the described specific orientation, or be constructed and operated in a specific orientation. Therefore, these descriptions should not be construed as limiting the invention. The terms "connection," "installation," "fixing," etc., should be interpreted broadly. For example, "connection" can be a fixed connection between multiple objects, a detachable connection between multiple objects, or an integral connection. It can be a direct connection between multiple objects or an indirect connection between multiple objects through an intermediate medium. For those skilled in the art, the specific meaning of the above terms in this invention can be understood based on the specific circumstances described above.

[0148] In the claims, description, and accompanying drawings of this invention, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In the claims, description, and accompanying drawings of this invention, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0149] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Those skilled in the art will recognize that the present invention can have various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method of determining a port arrival status of a vessel, characterized by, include: Acquire the navigation dynamic data of the target vessel, and determine the final position data of the target vessel's trajectory based on the navigation dynamic data; Based on the terminal position data of the flight path, multiple candidate ports within a preset navigation range are determined, and port data for each candidate port is obtained. Based on data from multiple ports, determine the dynamic proximity of the target vessel to each of the candidate ports; Convergence analysis is performed on multiple dynamic convergence values ​​to identify candidate ports that meet the convergence criteria from multiple port data as the target ports for the target vessels to dock. Based on the dynamic proximity and derived characteristics of the target port, the arrival confidence level is determined. Based on the arrival confidence level, the arrival status of the target vessel is determined.

2. The method for determining the arrival status of a ship according to claim 1, characterized in that, The step of acquiring the navigation dynamic data of the target vessel and determining the final position data of the target vessel's trajectory based on the navigation dynamic data includes: Acquire the navigation dynamic data of the target vessel, remove outliers and drift data from the navigation dynamic data, and organize the data into track data in chronological order; Based on the track data, determine the final position data of the target vessel's track.

3. The method for determining the arrival status of a ship according to claim 2, characterized in that, The step of determining the final position data of the target vessel's trajectory based on the trajectory data includes: The speed and heading data in the track data are smoothed, and the coordinate, time and distance units of all data in the track data are unified to generate standardized track data. Based on the standardized track data, the final position data of the target vessel's track is determined.

4. The method for determining the arrival status of a ship according to claim 1, characterized in that, Determining the dynamic proximity of the target vessel relative to each of the candidate ports based on multiple port data includes: Based on the ship's latitude and longitude and the candidate port coordinates, the spherical distance at consecutive time points is calculated using the semi-versus formula. The rate of change of distance is determined by the ratio of the difference in spherical distance at adjacent time points to the difference in time. Calculate the course deviation between the target vessel's current course and its ideal course toward the candidate port; Calculate the rate of change of speed based on the speed data at consecutive time points; By integrating the distance change rate, heading deviation, and speed change rate, a normalized dynamic convergence is obtained.

5. The method for determining the arrival status of a ship according to claim 1, characterized in that, The convergence analysis of multiple dynamic convergence values ​​determines candidate ports among multiple port data that meet the convergence criteria as target ports for the target vessels, including: Analyze the mean change trend and volatility of the dynamic convergence sequence corresponding to each candidate port; When the mean dynamic convergence of any candidate port continues to increase and the volatility continues to decrease and tends to stabilize, the candidate port is determined to meet the convergence condition. Candidate ports that meet the convergence conditions and whose convergence characteristics conform to preset conditions are identified as the target ports for the target vessels.

6. The method for determining the arrival status of a ship according to claim 1, characterized in that, The derived features include the mean, volatility, trend coefficient, speed change rate, heading stability, and distance change rate of the dynamic approach degree, as well as the target vessel's ship type, length, and gross tonnage data; the trend coefficient of the dynamic approach degree is the mean rate of change of the dynamic approach degree within a continuous time window; the heading stability is the standard deviation of the heading deviation within a continuous time window.

7. The method for determining the arrival status of a ship according to claim 1, characterized in that, Determining the arrival status of the target vessel based on the arrival confidence level includes: When the confidence level of arrival is greater than or equal to the first threshold, the target vessel is determined to be in a state of arrival. When the confidence level of arrival is greater than or equal to the second threshold and less than the first threshold, and there is a continuous upward trend, the target vessel is determined to be in the expected arrival state. When the confidence level of arrival is less than the second threshold, the target vessel is determined to be in a state of not having arrived. Wherein, the first threshold is greater than the second threshold, the first threshold is 0.8, and the second threshold is 0.

6.

8. A device for determining the arrival status of a ship, characterized in that, include: The first acquisition module is used to acquire the navigation dynamic data of the target vessel and determine the final position data of the target vessel's track based on the navigation dynamic data. The first determining module is used to determine multiple candidate ports within a preset navigation range based on the terminal position data of the flight path, and to obtain port data for each candidate port. The second determining module is used to determine the dynamic proximity of the target vessel to each of the candidate ports based on the port data from multiple sources. The third determination module is used to perform convergence analysis on multiple dynamic convergences and determine the candidate ports that meet the convergence conditions among multiple port data as the target ports for the target ships to dock. The fourth determining module is used to determine the arrival confidence level based on the dynamic proximity and derived features corresponding to the target port; The fifth determining module is used to determine the arrival status of the target vessel based on the arrival confidence level.

9. A device for determining the arrival status of a ship, characterized in that, include: processor; A memory containing programs or instructions, wherein the processor, when executing the programs or instructions in the memory, implements the steps of the method for determining the arrival status of a ship as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method for determining the ship's arrival status as described in any one of claims 1 to 7.

Images