A ship navigation mode extraction method and system

Abstract

The application discloses a ship navigation mode extraction method, which extracts and analyzes ship abnormal behaviors, ship frequent trajectories, ship periodic trajectories and ship trajectories under ship co-occurrence modes, can predict the probability of ship later running trajectories and abnormal behaviors, and can analyze the motion trend and motion law of the ship through a prediction structure, can perceive the situation, obtain more economic and safe navigation information, and improve the comprehensiveness and reliability of monitoring the ship navigation process.

Description

Technical Field

[0001] This application relates to the field of data mining technology, and more specifically, to a method and system for extracting ship navigation patterns. Background Technology

[0002] With the increasing prevalence of geographic information, the acquisition and computation capabilities of ship trajectory data are constantly evolving. The massive amounts of crowdsourced ship trajectory information collected by the Automatic Identification System (AIS) constitute a crucial component of ship trajectory data. Simultaneously, the vast amounts of situational intelligence data actively acquired by radar, surveillance, and satellite systems, along with the massive amounts of marine environmental data collected by numerous ocean observation programs such as Argo and Neptune, form a large, rich, and rapidly growing dataset. Motion pattern mining is a rapidly developing cutting-edge research topic in this field. Through trajectory pattern mining algorithms, valuable patterns can be discovered within large trajectory datasets, allowing for the analysis of ship motion trends and patterns to perceive the situation and obtain more economical and safer navigation information. However, current ship trajectory mining techniques involve large amounts of computational data, making it difficult to obtain accurate navigation trajectory data and comprehensively and effectively identify ship navigation trajectories. Summary of the Invention

[0003] To address at least one deficiency or improvement need in the prior art, the present invention provides a method and system for extracting ship navigation patterns, which can sense the situation, obtain more economical and safer navigation information, and improve the comprehensiveness and reliability of monitoring the ship navigation process.

[0004] To achieve the above objectives, according to a first aspect of the present invention, a method for extracting ship navigation patterns is provided, the method comprising:

[0005] AIS data is acquired through the Automatic Identification System (AIS) of ships, and the AIS data is preprocessed.

[0006] The DBSCAN algorithm is used to perform cluster analysis on ship trajectories to obtain ship clustering results.

[0007] Calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set a confidence interval based on the intra-class similarity, and identify ship trajectory patterns outside the confidence interval as abnormal ship behavior;

[0008] The trajectory feature centers of ship trajectory patterns with different intra-class similarities are calculated, the distance between the new ship trajectory and the trajectory feature centers of different ship trajectory patterns is calculated, and the category pattern of the new ship trajectory is determined to determine whether the new ship trajectory belongs to abnormal ship behavior.

[0009] Furthermore, in the above-mentioned method for extracting ship navigation patterns, the step of setting a confidence interval based on intra-class similarity and identifying ship trajectory patterns outside the confidence interval as abnormal ship behavior specifically includes:

[0010] Calculate the distance distribution function between ship trajectories, which conforms to a Gaussian normal distribution. Based on the Gaussian normal distribution, set a 95% confidence interval, and identify ship trajectory patterns outside the 95% confidence interval as abnormal ship behavior.

[0011] Furthermore, in the above-mentioned method for extracting ship navigation patterns, the step of calculating the trajectory feature centers of ship trajectory patterns with different intra-class similarities specifically includes:

[0012] Set category labels for different trajectory patterns in the ship clustering results;

[0013] The distance between all trajectories in different categories is calculated based on the DTW similarity measurement algorithm to generate a distance matrix;

[0014] Statistical analysis is performed on the distance matrix, and the trajectory represented by the row with the smallest sum of trajectory distances in the distance matrix is ​​taken as the trajectory feature center of this category.

[0015] Furthermore, the above-mentioned method for extracting ship navigation patterns also includes extracting frequent ship trajectories from the AIS data:

[0016] The preprocessed AIS data is used to construct a frequent trajectory tree and generate the time difference between adjacent nodes;

[0017] Generate trajectory paths based on the frequent trajectory tree, calculate the support and frequency of the trajectory paths, and save all trajectory paths with a support greater than the preset minimum support threshold and a trajectory frequency greater than the preset trajectory frequency to the frequent trajectory set.

[0018] Furthermore, the above-mentioned method for extracting ship navigation patterns also includes obtaining the ship's periodic trajectory through the AIS data:

[0019] The AIS data is used to detect dwell points and classify them.

[0020] Cluster analysis is performed on the classified stopping points to obtain the regions of interest for ship trajectories, and the periodic trajectory patterns in the regions of interest are found based on an Apriori-like algorithm.

[0021] Furthermore, the above-mentioned method for extracting ship navigation patterns also includes obtaining ship trajectories under ship co-occurrence patterns through the AIS data:

[0022] Ship data with temporal and spatial proximity relationships are extracted from the AIS data. The support and confidence scores between neighboring ships are calculated. Support and confidence thresholds are set, and ship trajectories that meet the support and confidence thresholds are selected as ship trajectories in the ship spatial co-occurrence mode.

[0023] Furthermore, in the above-mentioned ship navigation pattern extraction method, the probability of the occurrence of ship's subsequent operating trajectory and abnormal behavior is predicted based on the ship's abnormal behavior, frequent ship trajectory, periodic ship trajectory and ship trajectory in the ship co-occurrence pattern.

[0024] According to a second aspect of the present invention, a ship navigation pattern extraction system is also provided. This system includes a ship abnormal behavior extraction module, which further comprises a data processing submodule, a clustering analysis submodule, a ship abnormal behavior identification submodule, and a ship abnormal behavior discrimination submodule.

[0025] The data processing submodule is used to acquire AIS data through the Automatic Identification System (AIS) and preprocess the AIS data.

[0026] The clustering analysis submodule is used to perform clustering analysis on ship trajectories and obtain ship clustering results;

[0027] The ship abnormal behavior identification submodule is used to calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set a confidence interval based on the intra-class similarity, and identify ship trajectory patterns outside the confidence interval as ship abnormal behavior.

[0028] The ship abnormal behavior discrimination submodule is used to calculate the trajectory feature centers of ship trajectory patterns with different similarities within the same class, calculate the distance between the new ship trajectory and the trajectory feature centers of different ship trajectory patterns, and determine the category pattern of the new ship trajectory in order to determine whether the new ship trajectory belongs to ship abnormal behavior.

[0029] Furthermore, the aforementioned ship navigation pattern extraction system also includes a ship frequent trajectory extraction module, which is used to construct a frequent trajectory tree from the preprocessed AIS data and generate the time difference between adjacent nodes; generate trajectory paths based on the frequent trajectory tree, calculate the support and trajectory frequency of the trajectory paths, and save all trajectory paths with a support greater than a preset minimum support threshold and a trajectory frequency greater than a preset trajectory frequency to the frequent trajectory set.

[0030] Furthermore, the aforementioned ship navigation pattern extraction system also includes a ship periodic trajectory extraction module and a ship trajectory extraction module under ship co-occurrence patterns;

[0031] The ship periodic trajectory extraction module is used to detect stop points in the AIS data, classify the stop points, perform cluster analysis on the classified stop points to obtain the region of interest of the ship trajectory, and find the periodic trajectory pattern in the region of interest based on the Apriori-like algorithm.

[0032] The ship trajectory extraction module in the ship co-occurrence mode is used to extract ship data with temporal and spatial proximity relationships in the AIS data, calculate the support and confidence between neighboring ships, set support threshold and confidence threshold, and filter out ship trajectories that meet the support threshold and confidence threshold, which are regarded as the propagation trajectory in the ship spatial co-occurrence mode.

[0033] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0034] The ship navigation pattern extraction method provided by this invention can predict the probability of ship's subsequent operation trajectory and abnormal behavior by extracting and analyzing navigation patterns such as abnormal ship behavior, frequent ship trajectory, periodic ship trajectory and ship trajectory in co-occurrence patterns. By analyzing the ship's motion trend and motion law through prediction structure, it can perceive the situation and obtain more economical and safe navigation information, thereby improving the comprehensiveness and reliability of monitoring the ship's navigation process. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 A flowchart illustrating a method for extracting abnormal behavior patterns of ships provided in this application embodiment;

[0037] Figure 2 A schematic diagram illustrating the framework of a method for extracting abnormal ship behavior patterns provided in this application embodiment.

[0038] Figure 3 A schematic diagram illustrating the classification of abnormal behaviors provided in this application embodiment;

[0039] Figure 4 A framework diagram for frequent ship trajectory mining provided in the embodiments of this application;

[0040] Figure 5 A schematic diagram of dwell point detection is provided for the embodiments of this application;

[0041] Figure 6A schematic diagram of ship co-occurrence trajectories provided in the embodiments of this application;

[0042] Figure 7 A schematic diagram of the system framework for extracting ship navigation patterns. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0044] The terms "first," "second," "third," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0045] On the one hand, this application provides a method for extracting ship navigation patterns. Figure 1 This is a flowchart illustrating a method for extracting abnormal ship behavior patterns according to an embodiment of this application. Figure 2 Please refer to the schematic diagram of a method for extracting abnormal ship behavior patterns provided in this application embodiment. Figure 1 and Figure 2 The method includes the following steps:

[0046] (1) Obtain AIS data through the Automatic Identification System (AIS) and preprocess the AIS data;

[0047] Preferably, ship anomalies mainly include navigational anomalies and behavioral anomalies. Figure 3 This is a schematic diagram of abnormal behavior classification provided in the embodiments of this application. Abnormal navigation status includes whether the ship's position, trajectory, or distance is abnormal during navigation. Abnormal behavior characteristics include abnormal speed, heading, and navigation density.

[0048] Basic preprocessing of AIS data is performed, and abnormal and redundant data are detected and removed based on the latitude and longitude of the AIS data. A classification study of ship types is conducted, and dynamic behavioral characteristics of different ship types, such as speed, heading, and density, are statistically extracted. Normal state information for different characteristics is extracted, and abnormal behavior identification thresholds for different characteristics are constructed to assist in the identification of abnormal ship behavior.

[0049] (2) Cluster analysis of ship trajectories is performed based on the DBSCAN algorithm to obtain ship clustering results;

[0050] The DBSCAN algorithm is used to cluster ship trajectories, scanning existing clusters of stopping points. Clustering is performed when the data density is high. Generally, data with numerous periodic patterns inevitably contains a large number of high-density data clusters. First, an unprocessed point is extracted from the ship trajectory set. Then, the set of all current clusters is traversed, checking if the point belongs to a certain cluster. If such a cluster exists, the cluster center is updated. Otherwise, the point is considered a new cluster. Finally, the count of each cluster is checked to see if it meets the minimum count requirement. Based on the ship trajectory clustering results, features of different types of ship trajectories are extracted, ship behavior pattern information is mined, and different categories of ship pattern dictionaries are constructed.

[0051] (3) Calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set the confidence interval based on the intra-class similarity, and identify the ship trajectory patterns outside the confidence interval as abnormal ship behavior;

[0052] The intra-class similarity of different patterns in the ship trajectory clustering results is calculated based on the DTW algorithm. The calculation method is shown in the following formula.

[0053] d = d DTW (tr j' ,tr j )

[0054] j = 1, ..., n i j' = 1, ..., n i i = 1, ..., K

[0055] Where d represents the DTW distance between the two trajectories, tr j' ,tr j Let j and j' be the two trajectories to be tested, and j and j' be the sets of trajectory points in the two trajectories, respectively.

[0056] Calculate the distance distribution function between ship trajectories. The distance distribution function follows a Gaussian normal distribution. Fit its mean and variance, calculate the 95% confidence interval of the distance values, and identify ship trajectory patterns outside the 95% confidence interval as abnormal ship behavior. Set the maximum value of the 95% confidence interval of the distance values ​​as a similarity threshold as a criterion for identifying abnormal ship behavior. The calculation formula is as follows.

[0057]

[0058]

[0059] Where is the similarity distance of x, μ is the mean distance, and σ is the standard deviation. This represents the maximum similarity threshold for the i-th type of trajectory.

[0060] (4) Calculate the trajectory feature centers of the ship trajectory patterns with different intra-class similarity, calculate the distance between the new ship trajectory and the trajectory feature centers of different ship trajectory patterns, determine the category pattern of the new ship trajectory, and determine whether the new ship trajectory belongs to abnormal ship behavior.

[0061] Based on the clustering results, category labels are assigned to the trajectories of different clusters; the distance between all trajectories within different categories is calculated using the DTW similarity measurement algorithm, generating a distance matrix MC. i ,i=1,…,K; MC generated for different categories i Perform statistical analysis to determine the number of rows with the smallest distance sum, and set the trajectory with the smallest row sum as the representative feature center of this category.

[0062] Preferably, a nearest neighbor classifier is used to determine the category of the new trajectory, a similarity threshold is applied, and the trajectory is further judged to determine whether it is an abnormal trajectory.

[0063] The dataset is divided into a training set and a test set. The abnormal behavior recognition method of maritime trajectory is used to perform pattern mining and learning training on the training set, extract the feature centers of different patterns of trajectory, construct a pattern dictionary, and then perform abnormal behavior recognition on the actual maritime trajectory test set based on nearest neighbor classification.

[0064] The classification-based discrimination criteria are as follows:

[0065]

[0066] in This represents the similarity value between the j-th trajectory and the i-th feature center trajectory. This represents the maximum similarity threshold for the i-th type of trajectory determined based on the 95% confidence interval of trajectory similarity.

[0067] The ship navigation pattern extraction method provided in this application also includes a method for obtaining frequent ship trajectories. Figure 4 This document presents a framework diagram for mining frequent ship trajectories as provided in this application embodiment. Anomalies and redundant data are detected and removed based on the latitude and longitude of AIS data. The processed data is then dynamically constructed using a database scan, generating a frequent trajectory tree and time differences between adjacent nodes. If multiple ship trajectory paths pass through a certain point, this point should be merged into one item and stored in the item header table. The first few identical nodes of two trajectories starting from the head node are merged, and branches begin from different nodes. Time constraints are added to the mining of frequent trajectories, mining frequent paths or locations for different time periods. The time period constraints are determined based on user input. Considering practical application scenarios, locations are arranged in the order of the actual trajectories when constructing the trajectory tree, i.e., ship trajectories are arranged according to timestamps. A new, effective data structure is added to store the time length of the entire frequent path.

[0068] Define a dynamic array `temp[]` to store the candidate trajectory set during the trajectory mining process, and a dynamic array `frequency[]` to store the frequency of the trajectories. Initialize the frequent trajectory set `frequentSet[]` and the trajectory set `finalSet[]` that meets the frequency condition to empty. Based on the constructed frequent trajectory tree, first remove the support (l) from the tree. i For nodes whose support is less than min_supp (minimum support threshold), construct a subtree for the remaining items in the item header table. Merge identical trajectory sequences into a single path; these paths constitute a candidate path set S. Store set S in the temp[] array, and simultaneously store the frequencies of the candidate trajectories in the frequency[] array. Iterate through all paths in the temp[] array; if path S... i If the support is greater than or equal to min_supp, then the path S will be... i Store the data in a frequent trajectory set (frequentSet[]), and then iterate through the trajectory frequencies in frequency[]. route If support route If the value is within the mining frequency input by the user, the corresponding path will be stored in the finalSet[] array, thus obtaining the trajectory set that meets the frequency condition.

[0069] The ship navigation pattern extraction method provided in this application also includes a method for obtaining ship periodic trajectories. Figure 5This is a schematic diagram of dwell point detection provided in an embodiment of this application. Dwell point detection is performed on AIS data, and the dwell points are classified. Dwell points on ship tracks are divided into two categories: stationary dwell points and loitering dwell points. Stationary dwell points are those that remain at one point for a long time, with very close intervals between track points. Loitering dwell points indicate that the ship is circling around a certain point. Stationary dwell points can be identified using distance thresholds, speed, etc., while loitering dwell points require a step-by-step discussion:

[0070] ① For a trajectory point P, find a point P100 that is 100 meters away from it;

[0071] ② Calculate the time span between P and P100;

[0072] ③ If the time span exceeds the set threshold, the center point of these points that exceed the threshold will be identified as a wandering dwelling point.

[0073] Cluster analysis is performed on the set of stop points obtained in the previous step to identify the regions with the highest visit frequency as the regions of interest for the ships. Based on the DBSCAN clustering algorithm, existing stop point clusters are scanned, and clustering is performed when the data density is high. Generally, data with a large number of periodic patterns will inevitably contain a large number of high-density data clusters. First, an unprocessed point is extracted from the set of stop points. Then, the set of all current clusters is traversed, checking if the point belongs to a certain cluster. If such a cluster exists, the cluster center is updated. Otherwise, the point is considered a new cluster. Finally, it is checked whether the count of each cluster meets the minimum count requirement.

[0074] The algorithm is similar to Apriori, used to obtain periodically occurring trajectory patterns in all regions of interest. The periodic Apriori property is that every sub-pattern of a periodic pattern with period p is also a periodic pattern with period p. When a pattern exhibits periodicity, its sub-patterns also exhibit periodicity, and the period length remains constant. Based on this principle, the periodically occurring trajectory patterns in all regions of interest are identified.

[0075] The ship navigation pattern extraction method provided in this application also includes a method for obtaining ship trajectories under ship co-occurrence patterns. Figure 6 This is a schematic diagram of ship co-occurrence trajectories provided in an embodiment of this application. Ship data with temporal and spatial proximity relationships are extracted from AIS data, where spatial proximity is measured using Euclidean distance. Temporal proximity is measured using the difference in timestamps of ship position data instances.

[0076] The support and confidence scores among neighboring vessels are calculated. Support refers to the number of neighboring instances of an object in the dataset. When the number of neighboring instances exceeds a certain threshold, it is determined to be a spatiotemporal co-occurrence pattern. Confidence is the participation rate of instances in the spatiotemporal co-occurrence pattern, and its calculation formula is as follows:

[0077]

[0078] The confidence level condition is:

[0079] CPR (C k )>θ

[0080] Among them, C k ={f o ,...,f k-1} represents a k-ary candidate spatiotemporal co-occurrence pattern, f i Let E be the total set of spatiotemporal objects. Various combinations of ship objects constitute candidate spatiotemporal co-occurrence patterns. If the trajectory satisfies the support threshold and confidence threshold, it is a ship trajectory of the spatiotemporal co-occurrence pattern.

[0081] Based on the above-mentioned abnormal ship behavior, frequent ship trajectory, periodic ship trajectory and ship co-occurrence pattern, it is possible to predict the probability of ship's subsequent operation trajectory and abnormal behavior, analyze the ship's motion trend and motion law, so as to perceive the situation, obtain more economical and safer navigation information, and more comprehensively monitor the ship's navigation process.

[0082] On the other hand, this application also provides a ship navigation pattern extraction system. Figure 7 This is a schematic diagram of the framework of a ship navigation pattern extraction system. The system includes a ship abnormal behavior extraction module, which comprises a data processing submodule, a clustering analysis submodule, a ship abnormal behavior identification submodule, and a ship abnormal behavior discrimination submodule.

[0083] The data processing submodule is used to acquire AIS data through the Automatic Identification System (AIS) and preprocess the AIS data.

[0084] The clustering analysis submodule is used to perform clustering analysis on ship trajectories and obtain ship clustering results;

[0085] The ship abnormal behavior identification submodule is used to calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set a confidence interval based on the intra-class similarity, and identify ship trajectory patterns outside the confidence interval as ship abnormal behavior.

[0086] The ship abnormal behavior discrimination submodule is used to calculate the trajectory feature centers of ship trajectory patterns with different similarities within the same class, calculate the distance between the new ship trajectory and the trajectory feature centers of different ship trajectory patterns, and determine the category pattern of the new ship trajectory in order to determine whether the new ship trajectory belongs to ship abnormal behavior.

[0087] Preferably, the system also includes a ship frequent trajectory extraction module, which is used to construct a frequent trajectory tree from the preprocessed AIS data and generate the time difference between adjacent nodes; generate trajectory paths based on the frequent trajectory tree, calculate the support and trajectory frequency of the trajectory paths, and save all trajectory paths with a support greater than a preset minimum support threshold and a trajectory frequency greater than a preset trajectory frequency to the frequent trajectory set.

[0088] Preferably, the system further includes a ship periodic trajectory extraction module, used to detect dwell points in the AIS data, classify the dwell points, perform cluster analysis on the classified dwell points to obtain the region of interest of the ship trajectory, and find the periodic trajectory pattern in the region of interest based on an Apriori-like algorithm;

[0089] Preferably, the system further includes a ship trajectory extraction module under the ship co-occurrence mode, which is used to extract ship data in the AIS data that have temporal and spatial proximity relationships, calculate the support and confidence between neighboring ships, set support threshold and confidence threshold, and filter out ship trajectories that meet the support threshold and confidence threshold, which are regarded as propagation trajectories under the ship spatial co-occurrence mode.

[0090] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0091] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0092] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0093] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0094] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0095] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0096] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0097] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of embodiments of this disclosure upon considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

[0098] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and 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 for extracting ship navigation patterns, characterized in that, include: AIS data is acquired through the Automatic Identification System (AIS) of ships, and the AIS data is preprocessed. The DBSCAN algorithm is used to perform cluster analysis on ship trajectories to obtain ship clustering results. Calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set a confidence interval based on the intra-class similarity, and identify ship trajectory patterns outside the confidence interval as abnormal ship behavior; Calculate the trajectory feature centers of the ship trajectory patterns with different intra-class similarity, calculate the distance between the new ship trajectory and the trajectory feature centers of the different ship trajectory patterns, determine the category pattern of the new ship trajectory, and determine whether the new ship trajectory belongs to abnormal ship behavior. Specifically, the step of setting a confidence interval based on intra-class similarity and identifying ship trajectory patterns outside the confidence interval as abnormal ship behavior includes: Calculate the distance distribution function between ship trajectories, which conforms to a Gaussian normal distribution. Based on the Gaussian normal distribution, set a 95% confidence interval, and identify ship trajectory patterns outside the 95% confidence interval as abnormal ship behavior. The maximum value of the 95% confidence interval of the distance value is set as the maximum similarity threshold, calculated as follows: ； in, For similarity distance, The mean distance Standard deviation This represents the maximum similarity threshold for the i-th type of trajectory. This represents the total number of trajectories included in the i-th type of ship trajectory pattern; Specifically, calculating the trajectory feature centers of the ship trajectory patterns with different intra-class similarities includes: Set category labels for different trajectory patterns in the ship clustering results; The distance between all trajectories in different categories is calculated based on the DTW similarity measurement algorithm to generate a distance matrix; Statistical analysis is performed on the distance matrix, and the trajectory represented by the row with the smallest sum of trajectory distances in the distance matrix is ​​taken as the trajectory feature center of this category; The intra-class similarity of different patterns in the ship trajectory clustering results is calculated based on the DTW algorithm. The calculation method is shown in the following formula: ； in, These are two trajectories to be measured. and These are the sequence numbers of the two trajectories to be measured. This represents the DTW distance between the two trajectories to be measured.

2. The ship navigation pattern extraction method as described in claim 1, wherein, This also includes extracting frequent ship trajectories from the AIS data: The preprocessed AIS data is used to construct a frequent trajectory tree and generate the time difference between adjacent nodes; Generate trajectory paths based on the frequent trajectory tree, calculate the support and frequency of the trajectory paths, and save all trajectory paths with a support greater than the preset minimum support threshold and a trajectory frequency greater than the preset trajectory frequency to the frequent trajectory set.

3. The ship navigation pattern extraction method as described in claim 1, wherein, It also includes obtaining the ship's periodic trajectory through the AIS data: The AIS data is used to detect dwell points and classify them. Cluster analysis is performed on the classified stopping points to obtain the regions of interest for ship trajectories, and the periodic trajectory patterns in the regions of interest are found based on an Apriori-like algorithm.

4. The ship navigation pattern extraction method as described in claim 1, wherein, This also includes obtaining ship trajectories in ship co-occurrence patterns using the AIS data: Ship data with temporal and spatial proximity relationships are extracted from the AIS data. The support and confidence scores between neighboring ships are calculated. Support and confidence thresholds are set, and ship trajectories that meet the support and confidence thresholds are selected as ship trajectories in the ship spatial co-occurrence mode.

5. The method for extracting ship navigation patterns as described in any one of claims 1-4, wherein, Based on the ship's abnormal behavior, frequent ship trajectories, periodic ship trajectories, and ship trajectories in the co-occurrence pattern, the probability of the ship's subsequent operating trajectory and abnormal behavior is predicted.

6. A ship navigation pattern extraction system, characterized in that, This includes a ship abnormal behavior extraction module, which comprises a data processing submodule, a clustering analysis submodule, a ship abnormal behavior identification submodule, and a ship abnormal behavior discrimination submodule; wherein, The data processing submodule is used to acquire AIS data through the Automatic Identification System (AIS) and preprocess the AIS data. The clustering analysis submodule is used to perform clustering analysis on ship trajectories and obtain ship clustering results; The ship abnormal behavior identification submodule is used to calculate the intra-class similarity of different trajectory patterns in the ship clustering results, set a confidence interval based on the intra-class similarity, and identify ship trajectory patterns outside the confidence interval as ship abnormal behavior. The ship abnormal behavior discrimination submodule is used to calculate the trajectory feature centers of ship trajectory patterns with different similarities within the class, calculate the distance between the new ship trajectory and the trajectory feature centers of different ship trajectory patterns, determine the category pattern of the new ship trajectory, and thus determine whether the new ship trajectory belongs to ship abnormal behavior. Specifically, the step of setting a confidence interval based on intra-class similarity and identifying ship trajectory patterns outside the confidence interval as abnormal ship behavior includes: Calculate the distance distribution function between ship trajectories, which conforms to a Gaussian normal distribution. Based on the Gaussian normal distribution, set a 95% confidence interval, and identify ship trajectory patterns outside the 95% confidence interval as abnormal ship behavior. The maximum value of the 95% confidence interval of the distance value is set as the maximum similarity threshold, calculated as follows: ； in, For similarity distance, The mean distance Standard deviation This represents the maximum similarity threshold for the i-th type of trajectory. This represents the total number of trajectories included in the i-th type of ship trajectory pattern; Specifically, calculating the trajectory feature centers of the ship trajectory patterns with different intra-class similarities includes: Set category labels for different trajectory patterns in the ship clustering results; The distance between all trajectories in different categories is calculated based on the DTW similarity measurement algorithm to generate a distance matrix; Statistical analysis is performed on the distance matrix, and the trajectory represented by the row with the smallest sum of trajectory distances in the distance matrix is ​​taken as the trajectory feature center of this category; The intra-class similarity of different patterns in the ship trajectory clustering results is calculated based on the DTW algorithm. The calculation method is shown in the following formula: ； in, These are two trajectories to be measured. and These are the sequence numbers of the two trajectories to be measured. This represents the DTW distance between the two trajectories to be measured.

7. The ship navigation pattern extraction system as described in claim 6, wherein, It also includes a ship frequent trajectory extraction module, which is used to construct a frequent trajectory tree from the preprocessed AIS data and generate the time difference between adjacent nodes; generate trajectory paths based on the frequent trajectory tree, calculate the support and trajectory frequency of the trajectory paths, and save all trajectory paths with a support greater than the preset minimum support threshold and a trajectory frequency greater than the preset trajectory frequency to the frequent trajectory set.

8. The ship navigation pattern extraction system as described in claim 6, wherein, It also includes a ship periodic trajectory extraction module and a ship trajectory extraction module under ship co-occurrence mode; The ship periodic trajectory extraction module is used to detect stop points in the AIS data, classify the stop points, perform cluster analysis on the classified stop points to obtain the region of interest of the ship trajectory, and find the periodic trajectory pattern in the region of interest based on the Apriori-like algorithm. The ship trajectory extraction module in the ship co-occurrence mode is used to extract ship data with temporal and spatial proximity relationships in the AIS data, calculate the support and confidence between neighboring ships, set support threshold and confidence threshold, and filter out ship trajectories that meet the support threshold and confidence threshold, which are regarded as the propagation trajectory in the ship spatial co-occurrence mode.

Images