Research On Key Technologies Of Command And Control Behavior Modeling In CGF

The command and control modeling is an important application of CGF behavior modeling because it plays a core charator in combat.Aiming at the problems of modeling difficulty,weak flexibility and scalability and that the uncertainty and adversity of operations are not enough in the field of modeling of command and control behavior in CGF,the dissertation provides a general framework,method and tool support for modeling command and control behavior in CGF as the goal,focusing on the key technologies of command and control behavior modeling.The main contributions and innovations of the dissertation are as follows:(1)Propose a scalable and flexible framework of command and control behavior models.Through the research and analysis of the connotation and process of command and control in combat,the thesis made clear that the command and control behavior model of CGF must satisfy the functional requirements of reusablity and extensiblity,such as the collection and processing of intelligence,task-oriented decision-making and planning and collaborative capabilities.Aimed at the deficiency of the existing common CGF behavioral model framework,this paper constructs a command and control behavior model of CGF which contains five main modules,intelligence collection and processing,intelligent reasoning,general task management,task planning and coordination.It clarifies the functions and interrelations of each module,and points out that the highly efficient and flexible intelligent reasoning,general task management and adversarial task planning,which are three key technologies need to break through currently.(2)Design a rule-based intelligent reasoning model.To solve the reasoning inefficiency and model inflexibility of existing inference and decision-making methods in combat simulation applications,the paper designs a rule-based intelligent reasoning model.The model employs a rule-based knowledge base reasoning structure that allows each entity configure the knowledge base and the rules of combat flexibly,and reduce the difficulty of knowledge preparation and workload.To improve the flexibility of rule usage,the paper proposes a knowledge representation method based on facts,rules and actions and divides seven fact categories to allow users to reuse the rules at different stages of the triggering of generative rules.Combined with the corresponding inference control strategy and inference algorithm design,the model reduces the frequency of use of the knowledge base and reduces the number of rules checked per inference,which enhances the flexibility of use and improves the efficiency of reasoning.Finally,the dissertation clarifies the operating mechanism of the designed model in the simulation application and verifies the flexibility,high efficiency and scalability of the models based on an air-sea confrontation simulation.(3)Propose a generic task description method and general task management method based on Extended Hierarchical Task Network(E-HTN).The description and management of combat missions are important research topics in modeling the command and control of CGF behavior.In view of the problem of inadequate capacity to describe the relationship and environmental impact of operational tasks for stratified mission networks,based on the classic hierarchical task network(HTN),by introducing three concepts of mission-critical tasks,stages and exit conditions,the paper proposes an extension hierarchical task network(E-HTN)task description method,to enhance the ability to describe complex combat missions.Further,in order to solve the problems of poor flexibility,scalability and modeling difficulty,the paper proposes a task management method based on E-HTN,which provides CGF with a flexible combination of task definition and unified task management Service,improving the reusability and usability of the mission model.(4)Propose an adversarial HTN planning algorithm under partial observation condition:AHTN-R.In view of the current CGF command and control planning uncertainty of combat and confrontational considerations,the paper based on the E-HTN task description method,an improved confrontational hierarchical task planning method AHTN-R to address some of the dynamic environment confrontation planning under observable conditions.AHTN-R uses approximate state estimation methods.In order to solve the problem of task failure and re-planning in dynamic environment,AHTN-R adopts E-HTN-based failure task monitoring and priority-based task repair based on alternative solutions.The experimental results show that AHTN-R has good performance in the dynamic confrontation environment under partial observable conditions.(5)Design and develop the command and control behavior modeling tools.Based on the above research on the framework and key technologies,the dissertation designs and develops the command and control behavior modeling tool.First of all,it clarifies the functional requirements of the modeling tools for allegations and designs the overall structure and the operation flow of allegations of behavioral models.Secondly,it introduces the detailed design of the modules of task components development,combat rules editing and task editing modules.Finally,analyze the simulation application development and results under the background of offensive and defensive mission,the overall test of the work done in this paper.