Research On Modeling And Expression Of Scenario Knowledge Based On COLREGs

In recent years,with the continuous breakthrough of technologies such as sensors and artificial intelligence,vigorously developing intelligent ships has become an important means to promote the industrial restructuring of our country’s shipbuilding industry.The intelligent driving of ships includes the intelligent perception,intelligent computing,intelligent cognition and intelligent decision-making of the machine on the water traffic scenario.Among them,the perception,calculation and cognition of the machine to the scenario are the key to the intelligent decision-making of the machine,and it is also the basis for the realization of the intelligent driving of the ship.It is particularly important to carry out the research on the intelligent perception,calculation and cognition of the machine to the water traffic scenario for the realization of the intelligent navigation.Ship pilots can quickly integrate various scenario information in the process of ship navigation,capture key element information from complex navigation scenario information,establish the relationship between key element information,and consider the International Regulations for Preventing Collisions at Sea(COLREGs)constraints on ship behavior in different water traffic scenarios,thereby forming knowledge to guide ship behavior,but it is difficult for machines to achieve this,because machines lack the knowledge base to understand and judge the navigation scenario,so it cannot be completed.The transformation from sensor data to scenario knowledge and navigation situation knowledge.Therefore,how to convert various scenario knowledge expression of COLREGs into computer-identifiable and processable knowledge,so that machines can recognize and process COLREGs scenario similar to human brains,is of great significance for the research on ship intelligent navigation.This paper analyzes the multi-dimensional characteristics of COLREGs scenario knowledge,organizes and summarizes the COLREGs scenario knowledge system,constructs the COLREGs scenario knowledge model by means of resource description framework,description logic,sets and functions,etc.,and realizes the formal expression of COLREGs scenario knowledge.Ontology builds a COLREGs scenario knowledge cognition model,which realizes the machine’s cognition and reasoning of COLREGs scenario knowledge.Specific research results include the following:1.Analyzed COLREGs scenario knowledge elements and multi-dimensional characteristics,knowledge system and knowledge model.This paper analyzes the time,space and semantic characteristics of COLREGs scenario knowledge,summarizes the COLREGs scenario knowledge system,discusses the necessity of COLREGs scenario knowledge modeling and representation,and gives the theoretical framework of the COLREGs scenario knowledge model and knowledge expression method,for the construction of The conceptual model and cognitive model of scenario knowledge and the expression and reasoning of knowledge lay the theoretical foundation.2.Proposed a conceptual model and expression method of COLREGs scenario knowledge.The COLREGs scenario is abstractly expressed as a tuple model consisting of water traffic objects and their behaviors,and water traffic objects can be expressed as a triple(feature type,feature value,time)model of attribute elements and relation elements.On this basis,according to the dynamic change triplet of elements over time(element change type,change value,change time period),the behavior of water traffic objects is abstractly expressed as a triplet of attribute behavior and relationship behavior(element change type,semantic behavior,changing time period)model,and semantically express the behavior of scenario objects,which provides a good theoretical method basis for constructing the COLREGs scenario cognition model and realizing the intelligent cognition and computational reasoning of the COLREGs scenario by machines.3.Proposed a COLREGs scenario knowledge cognitive model and reasoning method.Based on the constructed conceptual model of COLREGs scenario knowledge,COLREGs scenario knowledge is extracted from sensor data and environmental data,and the ontology theory is introduced to construct a COLREGs scenario knowledge ontology cognitive model,revealing the inner evolution mechanism of COLREGs scenario,and realizing that machines can carry out the COLREGs scenario.Accurate cognition lays the foundation for making intelligent navigation behavior decisions in COLREGs scenarios.Finally,the case is verified by using the ship AIS data and environmental data of Xiamen Port.The experimental results show that the modeling and expression of the scenario in this paper can carry out the semantic cognitive reasoning of COLREGs scenario on the sensor data,and provide users with a ship-oriented intelligent navigation.The scenario semantic query mechanism.The COLREGs scenario knowledge model and expression method proposed in this paper provide new methods and ideas for intelligent navigation and ship-assisted decision-making.