Research On Unmanned Combat Vehicle Intelligent Planning Technology Based On Domain Knowledge Graph

The status of unmanned combat platforms(UCPs)in the future military struggle is becoming more and more important with the widespread use of artificial intelligence technology in the military.UCPs need higher autonomy and intelligence to adapt to various complex task types and battlefield environments.Introducing domain knowledge in the decision-making process not only enhances the UCP’s understanding of the situation,but also supports to make decisions on emergencies.This paper focus on the field of ground unmanned combat,proposes and constructs a knowledge graph model layer based on ontology,which is used for formal modeling domain knowledge.Secondly,it designs and constructs a task reasoning support system(TRSS)to deal with the uncertain information;Finally,this paper integrates the domain knowledge graph,TRSS and planning system to support the UCPs to generate a plan.The main work and contributions of this paper are as follows:(1)Proposed and constructed the ontology model of the knowledge graph in the ground unmanned combat field.This paper analyze the characteristics of ground unmanned combat domain knowledge,and construct the domain ontology based on OWL(Web Ontology Language).Secondly,the domain ontology constructed is probabilistic extended in the form of probability values in order to quantitatively describes the degree of uncertainty of domain knowledge.(2)Designed and constructed the MEBN-based TRSS.An uncertain knowledge reasoning framework TRSS is constructed in order to deal with the uncertain situation of the ground unmanned platform in the task planning execution.Based on domain knowledge and context-aware information,TRSS uses an accurate reasoning algorithm to make autonomous decisions and submit the decision results to the planning system in the form of tasks.Simulation experiments show that the constructed TRSS can make correct decisions on various emergencies encountered by unmanned combat platforms.(3)Designed and constructed an intelligent planning system based on domain knowledge graph.Firstly,this paper summarizes the existing knowledge-based intelligent planning system framework,analyzes the existing problems of the existing framework,and proposes the system framework in this paper.Then,this paper proposes a new OWL2 PDDL transformation algorithm to converted into the PDDL domain file required by the planner from the domain knowledge graph automatically;the planning system receives the decision results from the TRSS and adjusts the plan.For the ground unmanned combat scene,this paper carried out simulation experiments.Taking the unmanned combat vehicle to perform the fire force strike task as an example,set the emergency situation in the test process,and record the time costs for reasoning and planning.The timeliness of the intelligent planning system based on the domain knowledge graph designed in this paper is verified.