A water traffic ontology modeling method and model for holographic navigation scene graph
Through the water traffic ontology modeling method for holographic navigation scene diagrams, a semantic model of environment, rules and ship behavior is constructed, which solves the problem that traditional electronic channel diagrams cannot reflect object relationships, and realizes the multi-scale behavior of unmanned ships on traffic objects. cognitive and holographic semantic maps are applied to ship situation cognition and intelligent decision-making.
Patent Information
- Application Number
- CN202210596733.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-30
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-05-30
AI Technical Summary
Traditional electronic channel maps are difficult to reflect the hierarchical relationships and space-time relationships between objects, and cannot support querying the entire scene, especially the description of space-time attributes and direct querying causal relationships, resulting in the inability to realize machine understanding.
Using a water transportation ontology modeling method for holographic navigation scene diagrams, the environmental ontology model, regular ontology model and ship behavior ontology model are influenced by each other, and the scene ontology is constructed through semantic modeling and semantic extraction layers, using triple matching and deep learning. The algorithm recognizes scene categories, generates semantic descriptions of environmental elements, extracts international maritime collision avoidance rules, builds rules models, and builds behavior models based on ship activity data to form a holographic semantic map.
It realizes all-element, full-function and comprehensive semantic expression in water traffic scenarios, so that unmanned ships can perform multi-scale behavioral cognition of traffic objects, and form a holographic semantic map of behavioral scenes for unmanned ships to recognize them. cognition and intelligent decision-making on ships.
Smart Images

Figure CN114969928B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of scene modeling, and in particular relates to a water traffic ontology modeling method and model oriented to a holographic navigation scene graph. Background Art
[0002] The problem with traditional electronic navigation charts is that data queries only target a single element, making it difficult to reflect structural features such as hierarchical relationships between objects, and even more difficult to support queries on the entire scene. Secondly, traditional queries support queries on the temporal and spatial attributes of objects, but lack a description of temporal and spatial relationships and direct queries on the complete evolution of an object. Even if some queries can indirectly obtain historical information on a certain element, it is difficult to further query the causal relationships, action relationships, and other associations between different objects in that period. Current electronic navigation charts are difficult to reflect structural features such as object relationships, and cannot achieve machine understanding. Summary of the invention
[0003] The technical problem to be solved by the present invention is to provide a water traffic ontology modeling method and model oriented to a holographic navigation scene graph, which is used to enable unmanned ships to recognize the multi-scale behaviors of traffic objects.
[0004] The technical solution adopted by the present invention to solve the above technical problems is: a water traffic ontology modeling method for a holographic navigation scene graph, comprising the following steps:
[0005] S1: Deconstruct the water traffic ontology model for the holographic navigation scene graph into an ontology model that interacts with each other, including the environment ontology model, the rule ontology model, and the ship behavior ontology model;
[0006] S2: Construct an environment ontology model including a semantic modeling layer and a semantic extraction layer; the specific steps are:
[0007] S21: Construct the semantic modeling layer, and store the semantic information of the water transportation environment into the ontology knowledge base as a knowledge source after structured processing;
[0008] S22: Construct a semantic extraction layer to identify scene categories and generate semantic descriptions of environmental elements to achieve semantic understanding of the water traffic environment; based on the correlation between environmental scene categories and environmental elements, narrow the effective search range of unmanned ships when performing navigation tasks by identifying scene categories;
[0009] S23: driven by the navigation task, the automatic retransmission request ARQ query engine uses the triple matching method to retrieve the required environmental semantic information in the environmental scene ontology knowledge base to perform environmental scene recognition;
[0010] S3: Construct a rule ontology model by extracting elements and attributes including the International Regulations for Preventing Collisions at Sea;
[0011] S4: Taking ship behavior as the water traffic object, based on the environment scenario and rule scenario, using ship activity data, semantically express the activities of water traffic objects in the time and space range, and construct a ship behavior ontology model;
[0012] S5: Based on the environment ontology model, rule ontology model and ship behavior ontology model constructed in steps S2 to S4, information related to the environment, rules and ship behavior is obtained through corresponding perception means, and input into the corresponding ontology model to be instantiated as scene ontology information, forming a holographic semantic map of behavioral scenes within the scene domain for unmanned ships to recognize.
[0013] According to the above scheme, in step S21, the specific steps are:
[0014] S211: Based on the understanding of the formation process and constituent elements of water transportation environment scene semantics, the water transportation environment scene semantics are divided into different categories including natural environment and navigation infrastructure; the continuously distributed environmental field data are discretized with three-dimensional voxels as the basic unit, and the hydrological environment characteristics, meteorological environment characteristics and geographical environment characteristics of each basic unit are calculated according to the changes of various elements in the time dimension;
[0015] S212: Construct element semantic sets according to the characteristics and semantic features of different environmental scene elements, construct natural environment semantic sets used to describe the attributes and characteristics of water area meteorology, hydrology, and geographical environment, and construct navigation infrastructure functional semantic sets used to describe the attributes and functional characteristics of various navigation infrastructures;
[0016] S213: Realize the ontological expression of the semantics of the water transportation environment scene through ontology modeling; extract the concepts and attributes of the environment scene, analyze the relationship between the concepts, and use the ontology modeling tool to establish the ontology concept and reasoning rule library of the water transportation environment scene.
[0017] Furthermore, in the step S211, the natural environment includes meteorological environment, hydrological environment, and geographical environment characteristics; the meteorological environment includes wind speed, wind direction, visibility, air pressure, temperature and precipitation; the hydrological environment includes flow velocity, flow direction, wave height, main wave direction, wave period, tide time and tidal range; the geographical environment characteristics include topography and landforms; the navigation infrastructure refers to the key characteristics related to ship activities, including the navigation capacity of waterways, the distribution characteristics and berthing capacity of berths, the navigation function characteristics of navigation equipment and facilities, and the capacity and type characteristics of anchorages.
[0018] According to the above scheme, in step S22, the specific steps are:
[0019] S221: Identify environmental scene categories and generate semantic descriptions of environmental elements based on deep learning algorithms;
[0020] S222: Express the extracted semantic description in natural language and use ontology-based semantic parser to parse instances and attributes in the environment semantics;
[0021] S223: Match the environmental scene categories, instances and attributes with the environmental scene ontology knowledge base.
[0022] According to the above scheme, in step S3, the specific steps are:
[0023] S31: Analyze the types of water traffic rules from multiple aspects including procedural rules and empirical rules;
[0024] S32: Extract the elements of water traffic rules based on the ontology modeling method, abstract them into three elements: water traffic objects, object behaviors, and the relationship between objects or behaviors, as well as the attributes of these three elements; perform semantic extraction and semantic association, and construct a semantic model of the water traffic scene described by each rule;
[0025] S33: Perform collision avoidance role identification on ship objects with specified behaviors and attributes in each water traffic rule scenario; and perform semantic reasoning and semantic query on the operations that each ship should perform based on the ontology's reasoning and query mechanism, to form a rule knowledge base for smart ships with specific behaviors and attributes in specific scenarios, so as to realize the autonomous cognition of rules by smart ships and have the navigation awareness specified by the rules.
[0026] Furthermore, in step S31, the procedural rules include clear semantic descriptions of water traffic objects, behaviors, relationships and attributes; the empirical rules include "good seamanship" operations taken in real situations to improve the safety of water traffic objects, which are not specifically stated in the procedural rules but are necessary for water traffic rule scenarios and represent prior knowledge that smart ships possess.
[0027] According to the above scheme, in step S4, the specific steps are:
[0028] S41: constructing a motion state vector based on the activity data of the water transportation object, establishing a meta-behavior model of the water transportation object to describe the behavior semantic unit that characterizes the object's motion state, and transforming the object behavior from data to semantics;
[0029] S42: Based on the dimensional extended 9-intersection model DE-9IM, the topological relationship between the activity trajectory of water traffic objects and the environment in the water traffic scene is modeled and calculated to enrich the topological semantics of the object behavior; combining the trajectory characteristics and topological relationship of the object behavior, the traffic behavior model of the water traffic object is constructed to describe and express the meta-behavior of the water traffic object in a complete semantic space;
[0030] S43: Analyze ship behaviors at different scales from three different dimensions including time, space and semantics, divide traffic behaviors into three scales including microscopic behavior, mesoscopic behavior and macroscopic behavior, realize full-factor, full-function and all-round semantic expression in water traffic scenes, enable unmanned ships to recognize the multi-scale behaviors of their own ships and other traffic objects, and form a holographic semantic map of behavioral scenes within the scene domain for unmanned ships to recognize.
[0031] According to the above scheme, in the step S5, the corresponding sensing means include an environmental information collection system, a ship information collection system, and a rule information library; the environmental information collection system includes a waterway data collection system, a hydrological data collection system, and a meteorological data collection system; the waterway data collection system includes a waterway scanning data system, a high-resolution remote sensing satellite, and a high-resolution tilt camera; the hydrological data collection system includes a flow velocity and direction meter and a depth sounder; the meteorological data collection system includes a wind speed and direction meter; the ship information collection system includes a ship data collection system and a traffic flow data collection system; the ship data collection system includes a Beidou or GPS system, an inertial navigation system, a millimeter wave radar, Microwave radar, high-definition camera, infrared sensor, sound sensor; traffic flow data collection system includes AIS or LoRa system; rule information base includes international maritime or inland vessel collision avoidance regulations, port water navigation rules, other laws and regulations, good seamanship; instantiated scenario ontology information includes environment ontology, ship ontology, rule ontology; environment ontology includes water type, meteorological conditions, hydrological conditions; water type includes restricted waters and open waters; ship ontology includes ship type and ship status; ship type includes motor boats and fishing boats; ship status includes heading, speed, position, light type; rule ontology includes procedural rules and empirical rules.
[0032] A water traffic ontology model for a holographic navigation scene diagram includes an environment ontology model, a rule ontology model and a ship behavior ontology model; the environment ontology model is a natural environment semantic set used to describe the meteorological, hydrological and geographical environment attributes and characteristics of the water area, and a navigation infrastructure functional semantic set used to describe the attributes and functional characteristics of various navigation infrastructures; the environment ontology model includes an input and output layer, a semantic modeling layer and a semantic extraction layer connected in sequence; the input and output layer is used to input navigation mission requirements and output environmental semantic information; the semantic modeling layer is used to structure the semantic information of the water traffic environment and store it in an ontology knowledge base as a knowledge source; the semantic extraction layer is used to identify scene categories and generate semantic descriptions of environmental elements to achieve semantic understanding of the water traffic environment; the rule ontology model is to abstract the elements extracted from the water traffic rules into water traffic objects, behaviors, and relationship attributes, and perform semantic extraction and semantic association to construct a semantic model of the water traffic scene described by each rule; the ship behavior modeling is to use ship activity data to semantically express the activities of water traffic objects within the time and space range and construct a semantic model of the behavior scene.
[0033] A water traffic scene ontology model device for a holographic navigation scene graph comprises an environment ontology information processing module, a rule ontology information processing module, a behavior ontology information processing module, an ontology comprehensive expression module, a data interface, a user interface, a network interface and a power supply.
[0034] The beneficial effects of the present invention are:
[0035] 1. The water traffic ontology modeling method and model for the holographic navigation scene diagram of the present invention realizes full-factor, full-function and all-round semantic expression in the water traffic scene based on the scene semantic description of the environment, rules and ship behavior, so that the unmanned ship can meet the cognitive function of the multi-scale behavior of the traffic objects.
[0036] 2. The present invention forms a holographic navigation scene map within the scene domain for unmanned ships to recognize.
[0037] 3. The present invention can be applied to situational awareness and intelligent decision-making of ships. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 It is a schematic diagram of water traffic ontology modeling according to an embodiment of the present invention.
[0039] Figure 2 It is a schematic diagram of environmental ontology modeling according to an embodiment of the present invention.
[0040] Figure 3 It is a schematic diagram of rule ontology modeling according to an embodiment of the present invention.
[0041] Figure 4Schematic diagram of ship behavior ontology modeling according to an embodiment of the present invention.
[0042] Figure 5 It is an example diagram of the water traffic scene ontology model of an embodiment of the present invention.
[0043] Figure 6 It is a device diagram of the water traffic scene body according to an embodiment of the present invention. DETAILED DESCRIPTION
[0044] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0045] The present invention includes three parts: environment semantic modeling, rule semantic modeling and ship behavior modeling:
[0046] Environmental semantic modeling refers to the construction of a natural environment semantic set used to describe the attributes and characteristics of the water area's meteorological, hydrological, and geographical environments, and a navigation infrastructure functional semantic set used to describe the attributes and functional characteristics of various navigation infrastructures.
[0047] Rule semantic modeling refers to extracting elements from water traffic rules, abstracting them into water traffic objects, behaviors, and relationship attributes, and performing semantic extraction and semantic association to build a semantic model of the water traffic scene described by each rule;
[0048] Ship behavior modeling refers to the use of ship activity data to semantically express the activities of water traffic objects in time and space, and to construct a semantic model of the behavior scenario.
[0049] See also Figure 1 , a method according to an embodiment of the present invention comprises the following steps:
[0050] S1: Deconstruction of the water transportation scene ontology; see Figure 1 The water traffic ontology modeling is deconstructed into an interactive ontology model including the environment ontology model, the rule ontology model and the ship behavior ontology model.
[0051] S2: Environmental ontology modeling, including semantic modeling layer and semantic extraction layer; refer to Figure 2 .
[0052] The semantic modeling layer structures the semantic information of the traffic environment and stores it in the ontology knowledge base as a knowledge source; the semantic extraction layer identifies the scene category and generates the semantic description of the environmental elements to achieve semantic understanding of the traffic environment. The environmental scene cognition is driven by the navigation task and uses the triple matching method to retrieve the required environmental semantic information from the environmental scene ontology knowledge base.
[0053] At the semantic modeling layer, first, based on the understanding of the semantic formation process and constituent elements of the water transportation environment scene, it is divided into two categories: natural environment and navigation infrastructure. According to the characteristics of different environmental scene elements and their semantic features, the element semantic set is constructed. The natural environment mainly considers the meteorological environment (wind speed, wind direction, visibility, air pressure, temperature, precipitation, etc.), the hydrological environment (flow velocity, flow direction, wave height, main wave direction, wave period, tide time, tidal range, etc.), and the geographical environment characteristics (topography, landform, etc.). The navigation infrastructure mainly extracts the key features related to ship activities, including the navigation capacity of waterways, the distribution characteristics and berthing capacity of dock berths, the navigation function characteristics of navigation equipment and facilities, the capacity and type characteristics of anchorages, etc. Secondly, taking three-dimensional voxels as the basic unit, the continuously distributed environmental field data is discretized, and the changes of various elements in the time dimension are considered at the same time. The hydrological environment characteristics, meteorological environment characteristics and geographical environment characteristics are calculated for each basic unit. On this basis, a natural environment semantic set is constructed to describe the attributes and characteristics of the water area's meteorological, hydrological, and geographical environments, and a navigation infrastructure functional semantic set is constructed to describe the attributes and functional characteristics of various navigation infrastructures. Finally, the idea of ontology modeling is used to realize the ontology expression of the semantics of the water transportation environment scene, extract the concepts and attributes of the environment scene, analyze the relationship between the concepts, and use ontology modeling tools (such as OWL and SWRL) to establish the water transportation environment scene ontology concept and reasoning rule library.
[0054] At the semantic extraction layer, considering the correlation between environmental scene categories and environmental elements (such as port waters-berths), the recognition of scene categories can effectively narrow the search scope of unmanned ships when performing navigation tasks. Based on the deep learning algorithm, the environmental scene categories are identified; after the extracted semantic description is expressed in the form of natural language, the ontology-based semantic parser is used to parse out the instances and attributes in the environmental semantics, and finally the environmental scene categories, instances and attributes are matched with the environmental scene ontology knowledge base.
[0055] When performing navigation tasks, the unmanned ship is task-oriented and uses the automatic repeat request (ARQ) query engine to retrieve the required environmental semantic information from the ontology knowledge base by triple matching.
[0056] S3: Rule ontology modeling; reference Figure 3 By extracting the elements and attributes of rules such as the International Regulations for Preventing Collisions at Sea, a rule semantic model is constructed.
[0057] First, the types of water traffic rules are analyzed from two aspects: procedural rules and empirical rules. Procedural rules refer to clear semantic descriptions of water traffic objects, behaviors, relationships and attributes, including the International Regulations for Preventing Collisions at Sea, specific collision avoidance rules in some waters, navigation rules in port waters and related laws and regulations; while empirical rules refer to the "good seamanship" operations taken in real situations to improve the safety of water traffic objects. This rule is not specifically stated in the procedural rules but is indispensable in the water traffic rules scenario, representing the prior knowledge possessed by intelligent ships.
[0058] Secondly, based on the ontology modeling method, the elements of water traffic rules are extracted and abstracted into water traffic objects, object behaviors, the relationships between objects or behaviors, and the attributes of the first three. Semantic extraction and semantic association are performed to construct a semantic model of the water traffic scene described by each rule.
[0059] Finally, collision avoidance roles are identified for ship objects with specified behaviors and attributes in each water traffic rule scenario, and semantic reasoning and semantic query are performed on the operations that each ship should perform based on the ontology's reasoning and query mechanism, forming a rule knowledge base model for intelligent ships with specific behaviors and attributes in specific scenarios, enabling intelligent ships to autonomously recognize the rules and have the navigation awareness specified by the rules.
[0060] S4: Ship behavior ontology modeling; reference Figure 4 Taking ship behavior as the research object, based on environmental scenarios and rule scenarios, using ship activity data, the activities of water traffic objects in the time and space range are semantically expressed, and a semantic model of behavior scenarios is constructed.
[0061] Firstly, the motion state vector of water traffic objects is constructed based on their activity data, and a meta-behavior model of water traffic objects is established to describe the behavioral semantic units that characterize the motion state of the objects, thus realizing the transformation of object behavior from data to semantics.
[0062] Secondly, the topological relationship between the object's activity trajectory and the environment in the water traffic scene is modeled and calculated based on the Dimensional Extended 9-Intersection Model (DE-9IM) to achieve topological semantic enrichment of the object's behavior. Combining the trajectory characteristics and topological relationship of the object's behavior, a traffic behavior model for water traffic objects is constructed to describe and express the meta-behavior of water traffic objects in a complete semantic space.
[0063] Finally, the traffic behavior of ships is analyzed at different scales from three different dimensions: time, space, and semantics. The behavior is divided into three scales: activity (microscopic behavior), process (mesoscopic behavior), and event (macroscopic behavior). This realizes the full-factor, full-function, and all-round semantic expression in water traffic scenes, enabling unmanned ships to recognize the multi-scale behaviors of themselves and other traffic objects, and forming a holographic semantic map of behavioral scenes within the scene domain for unmanned ships to recognize.
[0064] S5: Instantiation of scene ontology information; see Figure 5 Based on the above three categories of ontology models, information about the environment, rules and ship behavior is obtained through corresponding perception means, and input into the corresponding ontology model to form a holographic semantic map of the behavior scene within the scene domain for unmanned ships to recognize.
[0065] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.
Claims
1. A water traffic ontology modeling method for holographic navigation scene graph, characterized by: The following steps are involved: S1: Deconstruct the water traffic ontology model for the holographic navigation scene graph into an ontology model that interacts with each other, including the environment ontology model, the rule ontology model, and the ship behavior ontology model; S2: Construct an environment ontology model including a semantic modeling layer and a semantic extraction layer; the specific steps are: S21: Construct a semantic modeling layer, and store the semantic information of the water transportation environment as a knowledge source in the ontology knowledge base after structural processing; the specific steps are: S211: Based on the understanding of the formation process and constituent elements of water transportation environment scene semantics, the water transportation environment scene semantics are divided into different categories including natural environment and navigation infrastructure; the continuously distributed environmental field data are discretized with three-dimensional voxels as the basic unit, and the hydrological environment characteristics, meteorological environment characteristics and geographical environment characteristics of each basic unit are calculated according to the changes of various elements in the time dimension; The natural environment includes meteorological environment, hydrological environment, and geographical environment characteristics; Meteorological conditions include wind speed, wind direction, visibility, air pressure, temperature and precipitation; The hydrological environment includes flow velocity, flow direction, wave height, main wave direction, wave period, tide time and tidal range; Geographical environmental features include topography and landforms; Navigation infrastructure refers to key features related to ship activities, including the navigability of waterways and routes, the distribution characteristics and berthing capacity of berths, the navigational function characteristics of navigational aids and facilities, and the capacity and type characteristics of anchorages; S212: Construct element semantic sets according to the characteristics and semantic features of different environmental scene elements, construct natural environment semantic sets used to describe the attributes and characteristics of water area meteorology, hydrology, and geographical environment, and construct navigation infrastructure functional semantic sets used to describe the attributes and functional characteristics of various navigation infrastructures; S213: Realize the ontology expression of the semantics of water transportation environment scene through ontology modeling; extract the concepts and attributes of the environment scene, analyze the relationship between the concepts, and use the ontology modeling tool to establish the ontology concept and reasoning rule library of the water transportation environment scene; S22: Construct a semantic extraction layer, identify scene categories and generate semantic descriptions of environmental elements to achieve semantic understanding of the water traffic environment; based on the correlation between environmental scene categories and environmental elements, narrow the effective search range of unmanned ships when performing navigation tasks by identifying scene categories; the specific steps are: S221: Identify environmental scene categories and generate semantic descriptions of environmental elements based on deep learning algorithms; S222: Express the extracted semantic description in natural language and use ontology-based semantic parser to parse instances and attributes in the environment semantics; S223: Corresponding the environment scene categories, instances and attributes to the environment scene ontology knowledge base; S23: Driven by the navigation task, the automatic retransmission request ARQ query engine uses the triple matching method to retrieve the required environmental semantic information in the environmental scene ontology knowledge base to perform environmental scene recognition; S3: By extracting the elements and attributes of the International Regulations for Preventing Collisions at Sea, a rule ontology model is constructed; the specific steps are: S31: Analyze the types of water traffic rules from multiple aspects including procedural rules and empirical rules; Procedural rules include clear semantic descriptions of water transportation objects, behaviors, relationships, and attributes; Empirical rules include "good seamanship" operations taken in real-life situations to improve the safety of water traffic objects. These are not specifically stated in procedural rules, but are necessary for water traffic rule scenarios and represent prior knowledge possessed by smart ships. S32: Extract the elements of water traffic rules based on the ontology modeling method, abstract them into three elements: water traffic objects, object behaviors, and the relationship between objects or behaviors, as well as the attributes of these three elements; perform semantic extraction and semantic association, and construct a semantic model of the water traffic scene described by each rule; S33: Perform collision avoidance role identification on the ship objects with specified behaviors and attributes in each water traffic rule scenario; and perform semantic reasoning and semantic query on the operations that each ship should perform based on the ontology reasoning and query mechanism, so as to form a rule knowledge base for intelligent ships with specific behaviors and attributes in specific scenarios, so as to realize the autonomous cognition of the rules by intelligent ships and have the navigation awareness specified by the rules; S4: Taking ship behavior as the water traffic object, based on the environmental scenario and rule scenario, using ship activity data, semantically express the activities of water traffic objects in the time and space range, and construct a ship behavior ontology model; the specific steps are: S41: constructing a motion state vector based on the activity data of the water transportation object, establishing a meta-behavior model of the water transportation object to describe the behavior semantic unit that characterizes the object's motion state, and transforming the object behavior from data to semantics; S42: Based on the dimensional extended 9-intersection model DE-9IM, the topological relationship between the activity trajectory of water traffic objects and the environment in the water traffic scene is modeled and calculated to enrich the topological semantics of the object behavior; combining the trajectory characteristics and topological relationship of the object behavior, the traffic behavior model of the water traffic object is constructed to describe and express the meta-behavior of the water traffic object in a complete semantic space; S43: Analyze ship behaviors at different scales from three different dimensions including time, space and semantics, and divide traffic behaviors into three scales including microscopic behavior, mesoscopic behavior and macroscopic behavior, so as to realize the semantic expression of all elements, functions and all aspects in water traffic scenes, so that unmanned ships can recognize the multi-scale behaviors of themselves and other traffic objects, and form a holographic semantic map of behavior scenes for unmanned ships to recognize within the scene domain; S5: Based on the environment ontology model, rule ontology model and ship behavior ontology model constructed in steps S2 to S4, information related to the environment, rules and ship behavior is obtained through corresponding perception means, and input into the corresponding ontology model to be instantiated as scene ontology information, forming a holographic semantic map of behavioral scenes within the scene domain for unmanned ships to recognize.
2. The method for water traffic ontology modeling based on holographic navigation scene graph according to claim 1 is characterized by: In the step S5, The corresponding sensing means include environmental information collection system, ship information collection system, and rule information database; The environmental information collection system includes the waterway data collection system, the hydrological data collection system and the meteorological data collection system; the waterway data collection system includes the waterway scanning data system, the high-resolution remote sensing satellite and the high-resolution tilt aerial camera; the hydrological data collection system includes the current velocity and direction meter and the depth sounder; the meteorological data collection system includes the wind speed and direction meter; The ship information collection system includes the ship data collection system and the traffic flow data collection system; the ship data collection system includes Beidou or GPS system, inertial navigation system, millimeter wave radar, microwave radar, high-definition camera, infrared sensor, sound sensor; the traffic flow data collection system includes AIS or LoRa system; The rules information base includes international regulations for preventing collisions between ships at sea or inland waters, navigation rules in port waters, other laws and regulations, and good seamanship; The instantiated scene ontology information includes environment ontology, ship ontology, and rule ontology; The environmental ontology includes water type, meteorological conditions, and hydrological conditions; water type includes restricted water and open water; The ship body includes the ship type and ship status; the ship type includes motor boats and fishing boats; the ship status includes heading, speed, position, and signal lights and types; The rule body includes procedural rules and empirical rules.
3. A water traffic ontology model system for holographic navigation scene graph, characterized by: Including environment ontology model, rule ontology model and ship behavior ontology model; The environmental ontology model is a natural environment semantic set used to describe the meteorological, hydrological, and geographical environmental attributes and characteristics of the water area, and a navigation infrastructure functional semantic set used to describe the attributes of various navigation infrastructure and their functional characteristics; the environmental ontology model includes an input-output layer, a semantic modeling layer, and a semantic extraction layer connected in sequence; the input-output layer is used to input navigation mission requirements and output environmental semantic information; the semantic modeling layer is used to structure the semantic information of the water transportation environment and store it in the ontology knowledge base as a knowledge source; the semantic extraction layer is used to identify scene categories and generate semantic descriptions of environmental elements to achieve semantic understanding of the water transportation environment; The rule ontology model abstracts the elements extracted from the water traffic rules into water traffic objects, behaviors, and relationship attributes, and performs semantic extraction and semantic association to construct a semantic model of the water traffic scene described by each rule. Ship behavior modeling is to use ship activity data to semantically express the activities of water traffic objects in time and space and construct a semantic model of behavior scenarios; Used to execute the water traffic ontology modeling method oriented to the holographic navigation scene graph as described in claim 1 or 2.
4. A water traffic scene ontology model device for a holographic navigation scene graph, characterized in that: It includes an environment ontology information processing module, a rule ontology information processing module, a behavior ontology information processing module, an ontology comprehensive expression module, a data interface, a user interface, a network interface and a power supply; Used to execute the water traffic ontology modeling method oriented to the holographic navigation scene graph as described in claim 1 or 2.
