Ship behavior knowledge graph construction method and device based on AIS data
By constructing a knowledge map of ship behavior and combining the time, space and semantic characteristics of AIS data, the problem of low efficiency in maritime departments in managing ship behavior is solved, and the convenience of automatic identification and efficient processing of ship behavior is achieved.
Patent Information
- Application Number
- CN202210131704.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-02-14
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-02-14
AI Technical Summary
Currently, there is a lack of mature ship behavior map construction methods, and it is difficult to effectively manage and use AIS data from the maritime department.
By obtaining the ship AIS data in the target area, a vector model of ship motion state is constructed, a ship behavior entity is identified, and a ship behavior ontology is constructed based on time, space and semantic characteristics, and a graph database is introduced to form a ship behavior knowledge graph.
It has improved the management efficiency and convenience of AIS data by the maritime department, and realized automatic identification and efficient processing of ship behavior.
Smart Images

Figure CN114564592B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of knowledge graph construction, and specifically relates to a method and device for constructing a ship behavior knowledge graph based on AIS data. Background Art
[0002] The outstanding performance of knowledge graphs in semantic expression, knowledge storage, and reasoning provides a technical foundation for data knowledge and intelligence. At the same time, knowledge graphs can also provide structured knowledge to users in a graphical manner. Currently, they have great applications in the fields of medicine, finance, agriculture, and education. However, there is currently no mature method for constructing ship behavior graphs. Summary of the invention
[0003] In view of the problems existing in the prior art, the present invention provides a method and device for constructing a ship behavior knowledge graph based on AIS data to realize the construction of a ship behavior knowledge graph.
[0004] To achieve the above object, the present invention provides a method for constructing a ship behavior knowledge graph based on AIS data, comprising the following steps:
[0005] Obtain AIS data of ships in the target area;
[0006] Construct a ship motion state vector model based on the acquired AIS data;
[0007] Identify ship behavior entities based on the ship motion state model;
[0008] Combine the characteristics of ship behavior in time, space and semantics to associate the identified entities and construct the ship behavior ontology;
[0009] Import the ship behavior ontology into the graph database to form a ship behavior knowledge graph.
[0010] Furthermore, the ship motion state vector is defined as:
[0011] Ship heading difference:
[0012] Ship acceleration:
[0013] Ship turning rate:
[0014] Stop speed threshold: v ε
[0015] Acceleration threshold: a ε
[0016] Unit heading difference threshold: δ
[0017] In the formula, c(t) represents the heading of the ship at time t, v(t) represents the speed of the ship at time t, and △t represents the time interval;
[0018] The obtained ship AIS data is calculated through the formula of the ship's motion state vector and the results are divided into different text descriptions according to human habits, including stay, movement, straight travel, turn, speed change, and constant speed.
[0019] Furthermore, the ship behavior entity recognition is specifically as follows: the entities are analyzed and extracted from the ship AIS data, namely, ship, behavior, time, track point, track, and then the environmental entity is established.
[0020] Furthermore, the characteristics of ship behavior at the time, space and semantic levels are obtained from the position, speed and heading of the ship AIS data. The ship behavior at the time level is affected by the AIS data transmission time interval and the ship speed. The representation of ship behavior at the spatial level is the area in the geographic space, that is, the area connected by different ship position points. The ship behavior at the semantic level is divided into the following three levels:
[0021] (1) Macro level: Based on the navigation behavior of ships in port waters, the macro behavior of port waters is divided into the following events: port entry-departure events, anchoring events, port entry events, berthing and unberthing events, and port departure events;
[0022] (2) Mesoscopic layer: In each navigation area, the ship engages in different processes and exhibits different mesoscopic behaviors as the topological relationship between the ship's motion trajectory and the navigation environment changes. Each event includes multiple processes, and the process behavior corresponds to the mesoscopic behavior at the time and space scale.
[0023] (3) Microscopic level: The motion state of the ship at the maneuvering level.
[0024] Furthermore, the ship behavior ontology includes ship class, behavior class, environment class, time class, trajectory class, and trajectory point class;
[0025] Determine the relationship between classes, store the classes and the relationship between classes in the graph database, and generate a knowledge graph of ship behavior.
[0026] Further, it is characterized in that
[0027] The ship class is the main object of behavioral cognition and the generator of trajectory and behavior. The ship's MMSI is used as the ship's unique identification code to uniquely determine the object of the behavior.
[0028] The behavior class is the core class in the behavior cognition ontology. The ship behavior is divided into four categories, which reflect different semantic types, including atomic behavior, topological behavior, traffic behavior and high-level semantic behavior. The atomic behavior class includes two subclasses: movement and stay. The movement behavior is used to describe the atomic behavior of different trajectory characteristics of ships. The topological behavior class, which consists of one or more actions, is the result of topological semantic enrichment based on the atomic behavior class. The topological behavior class represents the topological relationship between the ship trajectory and the environment, and the topological relationship corresponds one-to-one with the atomic trajectory of the ship. The traffic behavior class, as the basic semantic unit class of ship behavior, includes the activity class and is the key class for describing high-level semantic behavior. Traffic behavior describes the behavior reflected by the corresponding ship trajectory according to people's cognitive habits. The high-level semantic behavior class includes two categories: process and event. The subclasses of the high-level semantic behavior of ships are constructed and attributes are added according to different application scenarios.
[0029] Environmental categories refer to the different navigable areas designated;
[0030] The time class is a key element in recording ship behavior ontology, which is used to describe the start time, end time and duration of the behavior;
[0031] The trajectory class is the semantic abstraction of the ship trajectory data class, including atomic trajectory and complex trajectory. Trajectory and behavior are one-to-one correspondence and expression through the ontology description method;
[0032] The trajectory point class is one of the components of trajectory data, including the start and end points of the trajectory segment.
[0033] A device for constructing a ship behavior knowledge graph based on AIS data for implementing the above-mentioned ship behavior knowledge graph construction method based on AIS data comprises:
[0034] AIS data acquisition module, used to obtain AIS data of ships in the target area;
[0035] A vector model building module is used to build a ship motion state vector model based on the acquired AIS data;
[0036] A behavior entity recognition module is used to recognize ship behavior entities based on the ship motion state model;
[0037] The behavior ontology construction module is used to associate the identified entities with the characteristics of ship behavior in terms of time, space and semantics, and construct the ship behavior ontology;
[0038] The knowledge graph construction module is used to import the ship behavior ontology into the graph database to form a ship behavior knowledge graph.
[0039] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0040] The present invention provides a method and device for constructing a ship behavior knowledge graph based on AIS data, which can improve the management efficiency of AIS data by maritime departments and the convenience of using data. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 This is a flow chart of the method for constructing a ship behavior knowledge graph of the present invention;
[0042] Figure 2 A structural diagram of a device for constructing a ship behavior knowledge graph of the present invention;
[0043] Figure 3 This is a schematic diagram of ship behavior reasoning according to the present invention. DETAILED DESCRIPTION
[0044] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0045] Based on the actual situation in the industry, it is necessary to provide a method and device for constructing a ship behavior knowledge graph to identify the ship behavior in the target area and improve the maritime department's management efficiency of AIS data and the convenience of data use.
[0046] The present invention provides a method for constructing a ship behavior knowledge graph, comprising: acquiring ship AIS data in a target area, and constructing a ship motion state vector model according to the acquired AIS data; identifying ship behavior entities according to the ship motion state model; associating the identified entities with respect to the characteristics of the ship behavior in time, space and semantics; and constructing a ship behavior ontology, and then importing the ship behavior ontology into a graph database neo4j to form a ship behavior knowledge graph.
[0047] The characteristics of ship behavior are analyzed from the three dimensions of time, space and semantics, and the composition mechanism of ship behavior is analyzed; secondly, the ship behavior is divided into three types of behaviors: microscopic, mesoscopic and macroscopic based on the scale characteristics of the ship trajectory in time and space. Among them, the characteristics of ship behavior at the three levels of time, space and semantics can be obtained from the position, speed and heading analysis in AIS data. Ship behavior is affected by the AIS data transmission time interval and ship speed at the time level; the representation quantity at the spatial level is the area in the geographic space, that is, the area connected by different ship position points; at the semantic level, it can be divided into the following three levels according to people's habits:
[0048] (1) Macro level: According to the navigation behavior of ships in typical port waters, the macro behavior of port waters is divided into five events: port entry-departure events, anchoring events, port entry events, berthing and unberthing events, and port departure events.
[0049] (2) Mesoscopic layer: In each navigation area, the ship engages in different processes and exhibits different mesoscopic behaviors as the topological relationship between the ship's motion trajectory and the navigation environment changes. Therefore, each event includes multiple processes, and the process behavior corresponds to the mesoscopic behavior at the time and space scale.
[0050] (3) Microscopic level: The motion state of a ship at the maneuvering level, such as turning, accelerating, and stopping.
[0051] Secondly, considering the impact of the ship's navigation environment on the ship's behavior, it is necessary to combine the spatial characteristics of the ship's navigation environment, including point, line and surface areas, to model the interaction characteristics between the ship's trajectory motion unit and the navigation environment, and to enrich the ship's trajectory with topological semantics, so as to mine the implicit semantic information in the ship's trajectory and accurately identify the different behaviors of the ship.
[0052] The ship behavior ontology is constructed by using the ship time, space and semantic information obtained from the analysis of AIS data, including ship class, behavior class, environment class, time class, trajectory class, trajectory point class, and determine the relationship between classes. The data is stored in the graph database neo4j to generate a ship behavior map.
[0053] Furthermore, the ship behavior ontology includes the following contents:
[0054] (1) Ships
[0055] The ship is the subject of behavioral cognition and the generator of trajectory and behavior. The cognition of behavior requires the determination of the research object. The ship's MMSI (Maritime Mobile Satellite Identity) is the unique identification code of the ship, which can uniquely determine the object of the behavior.
[0056] (2) Behavioral
[0057] The behavior class is the core class in the behavior cognition ontology. The ship behavior is divided into four categories, which reflect different semantic types, including atomic behavior, topological behavior, traffic behavior and high-level semantic behavior.
[0058] The atomic behavior class includes two subclasses: motion and stay. The motion behavior class includes 6 subclasses: deceleration, acceleration, constant speed, right turn, left turn and direct sailing. The stay class indicates that the main engine of the ship stops rotating. The above two categories are used to describe the atomic behaviors of different trajectory characteristics of the ship. The topological behavior class is the result of topological semantic enrichment based on the atomic behavior class. It can represent the topological relationship between the ship trajectory and the environment. This topological relationship is used to represent the relationship between the four typical spatial objects in the geographic space, namely point and line, point and surface, line and line, and line and surface, and their space. As the basic semantic unit class of ship behavior, the traffic behavior class includes the activity class and is the key class for describing high-level semantic behavior. Traffic behavior can describe the behavior reflected by the corresponding ship trajectory according to people's cognitive habits. The high-level semantic behavior class includes two categories: process and event. The subclasses of the high-level semantic behavior of the ship are constructed and attributes are added according to different application scenarios.
[0059] (3) Environmental
[0060] The environment includes different navigation areas designated by the port authorities, such as anchorages, waterways, etc., traffic control areas and infrastructure, including waters in front of docks and buoys, etc. Some of these areas exist as physical entities, while others are designated by maritime authorities or port management departments and only displayed on nautical charts. Since different navigation environments have unique names or numbers and their respective location coordinates, the names or numbers can be used to conceptually define the navigation environment.
[0061] (4) Time
[0062] The time class is a key element in recording ship behavior ontology, which can be used to describe the start time, end time and duration of the behavior.
[0063] (5) Trajectory
[0064] The trajectory class is the semantic abstraction of the ship trajectory data class, including atomic trajectories and complex trajectories. Trajectories and behaviors are corresponded and expressed one by one through the ontology description method.
[0065] (6) Track point class
[0066] The trajectory point class is one of the components of trajectory data. Considering the key role of trajectory points in judging ship behavior, for example, when judging whether a moving trajectory segment and a closed area belong to entry / exit behavior, it is necessary to judge the topological relationship between the starting point and end point of the trajectory segment and the area respectively.
[0067] The ontology modeling tool protégé is used to construct the ship behavior ontology, and then it is imported into the graph database neo4j to form a ship behavior knowledge graph.
[0068] AIS data acquisition module, used to obtain AIS data of ships in the target area;
[0069] A vector model building module is used to build a ship motion state vector model based on the acquired AIS data;
[0070] A behavior entity recognition module is used to recognize ship behavior entities based on the ship motion state model;
[0071] The behavior ontology construction module is used to associate the identified entities with the characteristics of ship behavior in terms of time, space and semantics, and construct the ship behavior ontology;
[0072] The knowledge graph construction module is used to import the ship behavior ontology into the graph database to form a ship behavior knowledge graph.
[0073] The method and device for constructing a ship behavior knowledge graph of the present invention can realize automatic identification of ship behavior in a certain water environment, improve the processing efficiency of AIS data by the maritime department, and provide convenience for the maritime department to monitor ships.
[0074] Specifically, Figure 1 A method for constructing a ship behavior knowledge graph based on AIS data according to an embodiment of the present invention specifically comprises the following steps:
[0075] S100: AIS data acquisition. AIS data is obtained from the maritime authorities to extract ship track information. Ship track information contains a large amount of ship behavior information. For example, if a ship generates three consecutive track points at the same time, when these three track points are located on the same straight line or a curve, the ship is moving in a straight line or a curve; and when the three track points are located at the same position, it means that the ship is in a stop state. The spatial range of the track is different in the two cases.
[0076] S101: Ship motion state vector modeling. According to the different spatial characteristics of rules or infrastructure in the navigation scene, a spatial data model of the navigation scene is constructed, and the relationship between the ship trajectory and the entities in the scene is analyzed. In combination with the ship AIS data, a ship motion state vector model is constructed.
[0077] By analyzing the ship trajectory data, it is found that the trajectory data contains information such as location and time. Therefore, the ship's trajectory information can be abstracted into time latitude and space dimension. However, the above two dimensions are not enough to express the rich ship behavior. Then, a semantic model is constructed in combination with the ship motion characteristics to complete the ship motion state vector model. The ship motion vector is defined as:
[0078] Ship heading difference:
[0079] Ship acceleration:
[0080] Ship turning rate:
[0081] Stop speed threshold: v ε
[0082] Acceleration threshold: a ε
[0083] Unit heading difference threshold: δ
[0084] The semantic model of ship behavior refers to the calculation of the acquired AIS data through the formula of motion vector and the classification of the results according to human habits into text descriptions such as stay, move, go straight, turn, change speed, and constant speed.
[0085] S102: Ship behavior entity recognition: Analyze and extract entities such as ship, behavior, time, track point, track, etc. from AIS data, and establish environmental entities.
[0086] S103: Ship behavior entity relationship association. Reflect time indicates the time when the ship behavior occurs, before / after indicates the order of instant, hasBehavior indicates the relationship between the ship and the behavior, and hasTopo indicates the spatial topological relationship between the ship trajectory and the navigation environment.
[0087] S104: Use the ship behavior entity relationship in step S103 to associate the entities in step S102 to form a ship behavior ontology, store the ship behavior ontology in a graph database, and generate a ship behavior knowledge graph.
[0088] Figure 2 The structure diagram of the device for constructing the ship behavior knowledge graph based on AIS data of the present invention includes the following components:
[0089] The target area AIS data acquisition unit 200 is used to obtain AIS data from the maritime authority, extract the ship trajectory, analyze the trajectory data from the time and space dimensions, extract the ship motion vector and store it;
[0090] The ship behavior entity recognition unit 201 is used to store the ship behavior entity recognition program;
[0091] The ship behavior entity relationship association unit 202 is used to store the ship behavior entity relationship association rule program code;
[0092] The ship behavior reasoning unit 203 is used to call the unit programs 201-202 based on the pellet reasoning engine, and realize automatic reasoning of the ship behavior in combination with the custom SWRL reasoning rules. Two reasoning examples are given below:
[0093] Example 1: (Regularized process behavior) Anchor preparing process refers to the complex behavior of a ship from entering the anchorage to starting the anchoring stage, which is composed of a series of preparing to anchor activities. However, when reasoning about the high-level behavior of the ship, it is based on existing knowledge acquisition, and because the preparing to anchor activity is not easy to identify, the activity is not considered for knowledge reasoning. It is worth noting that the trajectory segment corresponding to the process behavior includes a series of motion sub-trajectories from cross_into_anchorage to the first anchor activity. These two behaviors can be obtained. Therefore, consider using two activities to constrain the ship's behavior, which can be expressed as follows using SWRL:
[0094] Anchor preparation process=cognition2:ship(?s)^cognition2:trajectory(?t)^cognition2:hasTraj(?s,?t)^cognition2:metaTraj(?stra)^cognition2:comprises(?t,?stra)^cognition2:Point(?p1)^cognition2:Point(?p2)^cognition2:hasBeginPoint(?stra,? p1)^cognition2:hasEndPoint(?stra,?p2)^cognition2:LA5(?stra,?p)^cognition2:anchorage(?p)^cognition2:hasSpeed(?p2,?x)^swrlb:lessThanOrEqual(?x,"0.5"^^xsd:float)->cognition2:hasBehavior(?s,cognition2:anchor_preparing)
[0095] Example 2: (Regularized event behavior) Unberth and Berth events represent all behaviors experienced by a ship in and around a berth, including approach pier activities, berthing processes, and leave pier activities, as shown below:
[0096] UnberthandBerthevent=cognition2:ship(?s)^cognition2:hasBehavior(?s,cognition2:Approach_pier)^co gnition2:reflectEnd(cognition2:Approach_pier,?i1)^cognition2:hasBehavior(?s,cognition2:berthing)^cognition2:reflectBegin(cognition2:berthing,?i2)^cognition2:reflectEnd(cognition2:berthing,?i3)^cognition2:hasBehavior(?s,cognition2:Leave_pier)^cognitio n2:reflectBegin(cognition2:Leave_pier,?i4)^time:before(?i1,?i2)^time:before(?i3,?i4)^time:Instant(?i1)^time:Instant(?i2)^time :Instant(?i0)^time:Instant(?i3)^time:Instant(?i4)^time:Instant(?i5)->cognition2:hasBehavior(?s,cognition2:berth_unberth_event)
[0097] Among them, the berthing process can be expressed as:
[0098] Berthing=cognition2:ship(?s)^cognition2:trajectory(?t)^cognition2:hasTraj(?s,?t)^cognition2:berth_activity(?t)->cognition2:hasBehavior(?s,cognition2:berthing)
[0099] The ship behavior knowledge graph 204 is a collection of outputs of the unit programs 201-203, forming a ship behavior knowledge graph.
[0100] Combined with neo4j, the ship behavior knowledge graph is constructed, and reasoning rules are constructed to realize the recognition and expression of ship behavior under certain time and space conditions. Figure 3The schematic diagram of the reasoning module of the ship behavior knowledge graph based on AIS data is shown, which is based on the SWRL reasoning rule. The function expression of the rule is represented by the class name or relationship name, and the variable is represented by a question mark "?" plus the variable name. By analyzing the complex behavior of ships in port waters, such as processes, events and other advanced behaviors, the behavior rules are defined with the help of SWRL, as shown in Table 1:
[0101] Table 1 Some SWRL inference rules for ship behaviors in different port waters
[0102]
[0103]
[0104] By reasoning about the behavior of the ship through the ontology, we can infer the behavior of the ship, the time when the behavior occurred, the location where the behavior occurred, and the object of the behavior.
[0105] Among them, Rule 1 means using the ship trajectory obtained from AIS data to represent the ship behavior;
[0106] Rules 2 and 4 respectively express the behavior of ships approaching and leaving the dock;
[0107] Rule 3 indicates that the ship’s track is determined using the AIS data ship position;
[0108] Rule 5 represents the ship's anchoring behavior;
[0109] Rule 6 refers to the behavior of ships passing through anchorages;
[0110] Rule 7 indicates that the object attributes cross the anchorage;
[0111] Rule 8 represents the conduct of navigation along the waterway;
[0112] Rule 9 represents time series;
[0113] Rule 10 represents the behavior of preparing for anchoring.
[0114] The above 10 rules demonstrate the implementation measures of the method and device for constructing a ship behavior knowledge graph based on AIS data in ship behavior identification and behavior reasoning. The method provided by the present invention can enable a computer to identify ship behavior, provide convenience for maritime authorities, and strengthen the management of ships.
[0115] In summary, the present invention provides a method and device for constructing a ship behavior knowledge graph, the method comprising: obtaining ship AIS data in a target area, and constructing a ship motion state vector model based on the acquired AIS data; identifying ship behavior entities based on the ship state vector model; associating ship behavior entities based on ship behavior characteristics; and constructing a ship behavior knowledge graph. The present invention also provides a device for constructing a ship behavior knowledge graph, which can identify ship behaviors in a target area, improve the efficiency of maritime departments in managing AIS data, and improve the convenience of using data.
[0116] It will be easily understood by those skilled in the art that the above are only preferred embodiments of the present invention and are 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 in the protection scope of the present invention.
Claims
1. A method for constructing a ship behavior knowledge graph based on AIS data, characterized in that: The following steps are involved: Obtain AIS data of ships in the target area; Construct a ship motion state vector model based on the acquired AIS data; Identify ship behavior entities based on the ship motion state model; Combine the characteristics of ship behavior in time, space and semantics to associate the identified entities and construct the ship behavior ontology; Import the ship behavior ontology into the graph database to form a ship behavior knowledge graph; Among them, the characteristics of ship behavior at the time, space and semantic levels are obtained from the position, speed and heading of the ship AIS data; the ship behavior at the time level is affected by the AIS data transmission time interval and the ship speed; the representation of ship behavior at the spatial level is the area in the geographic space, that is, the area connected by different ship position points; ship behavior at the semantic level is divided into the following three levels: (1) Macro level: Based on the navigation behavior of ships in port waters, the macro behavior of port waters is divided into the following events: port entry-departure events, anchoring events, port entry events, berthing and unberthing events, and port departure events; (2) Mesoscopic layer: In each navigation area, the ship engages in different processes and exhibits different mesoscopic behaviors as the topological relationship between the ship's motion trajectory and the navigation environment changes. Each event includes multiple processes, and the process behavior corresponds to the mesoscopic behavior at the time and space scale. (3) Microscopic level: the motion state of the ship at the maneuvering level; The ship behavior ontology includes ship class, behavior class, environment class, time class, trajectory class, and trajectory point class. The inter-class relationship is determined, and the classes and inter-class relationships are stored in the graph database to generate a ship behavior knowledge graph. The ship class is the subject object of behavior cognition and the generator of trajectory and behavior. The MMSI of the ship is the unique identification code of the ship, which uniquely determines the object of the behavior. The behavior class is the core class in the behavior cognition ontology. The ship behavior is divided into four categories, which reflect different semantic types, including atomic behavior, topological behavior, traffic behavior and advanced semantic behavior. The atomic behavior class includes two subclasses: motion and stay. The motion behavior is used to describe the atomic behavior of different trajectory characteristics of ships. The topological behavior class is the result of topological semantic enrichment based on the atomic behavior class, which represents the topological relationship between the ship trajectory and the environment. This topological relationship is used to represent the relationship between the four typical spatial objects of point and line, point and surface, line and line, and line and surface in the geographic space and their space. The traffic behavior class is the basic semantic unit class of ship behavior, including the activity class, which is the key class for describing advanced semantic behavior. Traffic behavior describes the behavior reflected by the corresponding ship trajectory according to people's cognitive habits. The advanced semantic behavior class includes two categories: process and event. The subclasses of the ship's advanced semantic behavior are constructed and attributes are added according to different application scenarios. Environmental categories refer to the different navigable areas designated; The time class is a key element in recording ship behavior ontology, which is used to describe the start time, end time and duration of the behavior; The trajectory class is the semantic abstraction of the ship trajectory data class, including atomic trajectory and complex trajectory. Trajectory and behavior are one-to-one correspondence and expression through the ontology description method; The trajectory point class is one of the components of trajectory data, including the start and end points of the trajectory segment.
2. The method for constructing a ship behavior knowledge graph based on AIS data according to claim 1 is characterized in that: The ship motion state vector is defined as: Ship heading difference: Ship acceleration: Ship turning rate: Stop speed threshold: v ε Acceleration threshold: a ε Unit heading difference threshold: δ In the formula, c(t) represents the heading of the ship at time t, v(t) represents the speed of the ship at time t, and △t represents the time interval; The obtained ship AIS data is calculated through the formula of the ship's motion state vector and the results are divided into different text descriptions according to human habits, including stay, movement, straight travel, turn, speed change, and constant speed.
3. The method for constructing a ship behavior knowledge graph based on AIS data according to claim 1, characterized in that: The specific process of ship behavior entity recognition is as follows: analyzing and extracting entities from ship AIS data, namely, ship, behavior, time, track point, track, and establish environmental entities.
4. A device for constructing a ship behavior knowledge graph based on AIS data for implementing the method for constructing a ship behavior knowledge graph based on AIS data as described in any one of claims 1 to 3, characterized in that: include: AIS data acquisition module, used to obtain AIS data of ships in the target area; A vector model building module is used to build a ship motion state vector model based on the acquired AIS data; A behavior entity recognition module is used to recognize ship behavior entities based on the ship motion state model; The behavior ontology construction module is used to associate the identified entities with the characteristics of ship behavior in terms of time, space and semantics, and construct the ship behavior ontology; The knowledge graph construction module is used to import the ship behavior ontology into the graph database to form a ship behavior knowledge graph.
