Research On Scalable Distributed Organizational Structure And Dynamic Reorganization Mechanisms Of Unmanned Systems

With the increasing complexity of unmanned system application scenarios and tasks,a single unmanned system cannot satisfy the demand.Complex missions require multiple unmanned systems to collaborate to form clusters and work cooperatively to accomplish the tasks.Currently,unmanned systems with a centralized architecture will experience a single point of failure.Once the central node fails,the whole unmanned system will not work typically.In addition,the unmanned system lacks scalability and is challenging to cope with the dynamic change of the number of unmanned systems.The unmanned system cannot independently adjust and support the loss or addition of some unmanned systems.In the traditional unmanned system task replanning,the global task replanning will be carried out,which is difficult to improve the efficiency of task replanning.This thesis studies the distributed scalable architecture of the unmanned system,the dynamic self-organization mechanism,and the task replanning method in the dynamic reorganization mechanism.This thesis proposes a flexible three-layer scalable distributed architecture and a dynamic self-organization mechanism based on this architecture.A dynamic replanning method is proposed with the dynamic self-organization mechanism.In response to the single point of failure and non-scalability problems that characterize unmanned systems,a flexible three-layer scalable distributed architecture model of the "1+C+X" mode is proposed.Unmanned systems with distributed architecture will not have single-point paralysis.The automatic reconfiguration of the control structure can be supported through flexible three-layer scalable architecture,making the unmanned system scalable.Furthermore,based on the flexible three-layer scalable distributed architecture,a dynamic self-organization mechanism is proposed to ensure that the unmanned system in the unmanned system can be dynamically increased or decreased.Finally,aiming at the task replanning problem caused by the increase or decrease of unmanned systems in the unmanned systems,and efficient replanning algorithm is proposed based on a dynamic self-organization mechanism and contract network protocol algorithm.Through this algorithm,the unmanned system can carry out local task replanning according to the increase or decrease of the unmanned system.It can realize the rapid response to the situation change,generate a new planning scheme in as little time as possible,and significantly improve the work efficiency of the unmanned system.The unmanned surface vehicle(USV)is the unmanned system entity,taking three typical unmanned system task modes of formation navigation.In this thesis,the task scalability and reconfigurability of unmanned systems are studied by using the USV group cooperative control simulation platform based on Vega software and take three typical mission modes as examples:formation navigation,regional patrol,and escort.Simulation experiments verify the rationality of the flexible three-layer scalable distributed architecture model of "1+C+X" mode,verify the effectiveness of the dynamic self-organization mechanism,and realize the task replanning of unmanned systems.This makes the unmanned system with the USV as the entity of unmanned system scalable and can meet the task requirements.