Research On Bayesian Network Threat Assessment And Decision Method For UUV Swarm Attack And Defense Mission

The issue of territorial sea security has increasingly become the focus of attention in the maritime field.In view of the advantages of the Unmanned Underwater Vehicle(UUV),the joint operation of UUV swarm has become the normal choice for the task of ensuring the safety of territorial waters.This paper takes the blue UUV attack and the red UUV defense as the study background,and conducts research on UUVs threat assessment and decision-making methods to improve the autonomous ability of UUV group operations.First,studying the UUV characteristics and architecture.This studying part Clarifies the UUV composition characteristics,the UUV components,and analyzes the control information transmission process in the UUV control system.Based on the traditional architecture model,a UUVs architecture model is proposed,which clarifies the timing logic of threat assessment and decision-making methods during the execution of UUVs tasks.Secondly,studying on UUVs threat assessment methods.Studying on the Bayesian network method that gives the mathematical description of the Bayesian network and gives the recurrence formula of the Bayesian network.The threat assessment is divided into two parts: intention assessment and security assessment.The intention assessment is used to evaluate the target’s threat intention and threat level to the red base,and the security assessment is used to evaluate the working status of UUV.Based on a single UUV,the realization methods of UUV intention assessment and UUV safety assessment are studied.Based on the UUV swarm,the implementation of UUVs intention assessment and UUVs safety assessment is studied.Third,studying on UUVs decision-making methods based on the threat assessment results.The research on UUVs decision-making methods is divided into two parts: UUVs dominant collaborative decision-making and UUVs task allocation.Aiming at the problem of UUVs dominant collaborative decision-making,the study is based on the static game model of incomplete information in game theory.Explain the basic concepts and mathematical description of the game theory method.Since the goal of the blue party is unknown,based on the static game model with incomplete information,a Bayesian Nash equilibrium calculation method for the content of this topic is given to balance the split point and the UUV group intention.The comparison of evaluations serves as the basis for UUVs to dominate collaborative decision-making.Aiming at the problem of UUVs task allocation,the research is based on the distributed auction algorithm.The basic concept and implementation process of the distributed auction algorithm are described.The cost function is established for UUV safety assessment,and the time constraint is studied.The UUV reachability analysis under time-limited conditions is given,and the task allocation model is constructed based on the UUV group.Finally,the issue of UUV sports strategy is studied.Based on the idea of artificial potential field method and based on the repulsion force model,the red and blue UUV movement strategies are formulated.According to the proposed motion strategies of various units of red UUVs,simulation verification is carried out.Aiming at the methods proposed above,a comprehensive simulation experiment is performed under a given verification scenario to verify the feasibility and rationality of the proposed UUVs threat assessment and decision-making method.